Discover Computer Vision Projects Efficiently

AI Developer Needed – Multi-Agent Real Estate Acquisition Platform

We are looking for an experienced AI developer to build a multi-agent AI platform that automates commercial real estate acquisition across the United States.

The platform should include AI agents that can:

* Find hard-corner, freeway exit, and high-potential commercial properties using GIS, satellite imagery, parcel, and traffic data.
* Analyze properties based on visibility, traffic counts, zoning, accessibility, and development potential.
* Identify property owners and gather contact information (phone, email, mailing address, business entities).
* Automatically send personalized email outreach and follow-up campaigns.
* Integrate an AI voice agent to call property owners, qualify leads, and schedule appointments.
* Store and rank properties in a searchable database with interactive maps and dashboards.

Preferred Skills

* AI Agents (LangGraph, CrewAI, AutoGen, OpenAI Agents, etc.)
* Python
* GIS / Geospatial Development (PostGIS, ArcGIS, Google Maps, Mapbox)
* Computer Vision & Satellite Imagery
* API Integrations
* Web Scraping & Data Enrichment
* Machine Learning / Predictive Analytics
* CRM, Email, and AI Voice integrations (Twilio, Bland AI, Vapi, Retell AI, etc.)
* AWS, Azure, or Google Cloud

Please include:

* Similar AI or real estate projects you've built
* Your recommended tech stack and architecture
* Estimated timeline and budget
* Whether you've built multi-agent AI systems before

Bonus: Experience with commercial real estate, property data, GIS, or land acquisition platforms.

We are looking for an experienced Computer Vision Engineer to urgently audit a drive-by dataset collected for a vegetation and roadside-monitoring project.

This is the first phase of a larger computer-vision project. We are looking for someone who can start immediately and, if the audit is completed successfully, continue with us on a long-term basis for annotation, model fine-tuning, evaluation and implementation.

Project background

The dataset contains approximately 1 hour and 40 minutes of drive-by recordings.

The available data may include:

- RGB images or video frames;
- frame and chunk identifiers;
- GPS coordinates;
- timestamps;
- route information;
- LiDAR data or LiDAR-derived height statistics;
- YOLO-World object-detection output;
- SegFormer vegetation-segmentation output;
- JSON and GeoJSON files;
- previously identified false positives.

The purpose of the audit is to determine:

- what data is actually available;
- how much of the data is usable and unique;
- which classes are sufficiently represented;
- whether the existing model outputs are useful;
- whether GPS and LiDAR can be linked reliably;
- what is required for a representative annotation and model-training pilot.

Scope of the audit

1. File and dataset inventory

Identify and document:

- all available files and formats;
- total number of videos, chunks and frames;
- available RGB, GPS, LiDAR, JSON and GeoJSON data;
- missing, corrupted or unreadable files;
- links between frames, chunks, routes and sensor data.

2. Image-quality assessment

Assess the dataset for:

- motion blur;
- poor focus;
- compression issues;
- shadows and backlighting;
- distance from relevant objects and vegetation;
- visibility of roadsides, vegetation and fixed obstacles;
- unusable or low-quality frames.

This may be performed using scripts combined with a representative manual review. We do not expect every frame to be reviewed manually.

3. Duplicate and redundancy analysis

Determine:

- how many identical or near-identical frames are present;
- how much redundancy exists within and between chunks;
- whether locations or route sections were recorded repeatedly;
- how the dataset can be reduced to representative frames;
- which frame-selection method should be used instead of automatically selecting every fifth frame.

4. Location-based dataset structure

Review:

- GPS coordinates and timestamps;
- unique locations and routes;
- repeated locations and route overlap;
- whether training, validation and test data can be separated by location;
- how data leakage between near-identical frames and locations can be prevented.

5. Preliminary class distribution

Assess whether the data contains usable examples of the following categories.

Fixed mowing obstacles

- tree trunks;
- posts, bollards and lamp posts;
- traffic signs;
- fences, gates and guardrails;
- benches and picnic tables;
- utility cabinets, mailboxes and hydrants;
- large rocks or boulders;
- culverts and bridge structures;
- other permanent mowing obstacles.

People, animals and moving vehicles are outside the target scope.

Vegetation

- short or mown grass;
- tall roadside vegetation;
- reed or watercourse vegetation;
- woody vegetation;
- cut vegetation or temporary vegetation deposits;
- bare or recently mown ground;
- other or uncertain vegetation.

The audit should indicate:

- which classes occur;
- estimated example counts per class;
- which classes have insufficient data;
- which classes should initially be grouped;
- which classes are not visually distinguishable with the available image quality.

6. Existing model-output review

Review a representative sample of the available:

- YOLO-World detections;
- SegFormer segmentations;
- confidence scores;
- false positives;
- missed objects or vegetation regions.

Determine whether the existing outputs are suitable for automatic pre-annotation and whether correcting them is likely to be faster than annotating from scratch.

No model fine-tuning is required during this audit unless separately agreed.

7. GPS and LiDAR assessment

Determine:

- whether timestamps are consistent;
- whether LiDAR data can be linked to the correct frames, chunks and locations;
- whether the available calibration and sensor information is sufficient;
- whether LiDAR can support vegetation-height analysis;
- whether reliable use is possible at pixel, frame, chunk or location level;
- which technical limitations or missing information exist.

The audit does not include development of a complete RGB–LiDAR fusion pipeline unless separately agreed.

8. Pilot recommendation

Based on the audit, provide a clear recommendation for the next pilot phase, including:

- recommended number of pilot frames;
- frame-selection method;
- proposed final class structure;
- proposed annotation approach;
- location-based training, validation and test split;
- estimated annotation effort;
- expected technical risks;
- additional data that may be required;
- estimated scope and cost of the next phase.

Required deliverables

The final audit must include:

1. documented dataset inventory;
2. counts of available, usable, unusable and duplicate data;
3. overview of unique routes and locations;
4. image-quality findings;
5. preliminary class distribution;
6. review of existing YOLO-World and SegFormer outputs;
7. assessment of GPS and LiDAR usability;
8. recommended frame-selection strategy;
9. recommended pilot dataset;
10. proposed train, validation and test structure;
11. estimated annotation effort;
12. technical risks and missing information;
13. recommended next steps;
14. concise written audit report;
15. all scripts used to produce the audit results.

All findings must be reproducible and based on the supplied data.

Data-integrity requirement

We will independently verify:

- file and frame counts;
- duplicate percentages;
- class counts;
- GPS and route findings;
- model-output observations;
- samples used in the report;
- scripts and generated outputs.

Do not present invented, synthetic or estimated findings as measured results.

Every estimate must be clearly labelled as an estimate and must include the method, sample or assumption on which it is based.

Fabricated data, unverifiable metrics, manipulated screenshots or non-reproducible conclusions will result in immediate rejection of the work and termination of the contract.

Required experience

You should have demonstrable experience with:

- computer-vision datasets;
- Python;
- OpenCV;
- YOLO or comparable object-detection models;
- semantic segmentation;
- SegFormer or comparable segmentation models;
- dataset-quality audits;
- duplicate and near-duplicate image detection;
- train, validation and test dataset design;
- GPS or geospatial data;
- JSON and GeoJSON.

Experience with LiDAR, camera–LiDAR calibration or sensor fusion is strongly preferred.

Urgency and availability

This is an urgent assignment.

Please apply only if you can:

- start immediately or very soon;
- work directly with the supplied raw data;
- communicate progress clearly;
- provide reproducible findings;
- state your expected completion date before starting.

Long-term opportunity

This audit is the first part of a broader project.

A successful audit may lead to further work involving:

- annotation strategy and quality control;
- object-detection fine-tuning;
- vegetation segmentation;
- LiDAR-assisted height analysis;
- false-positive reduction;
- GPS and GeoJSON output;
- evaluation on unseen locations;
- production-pipeline development.

We are specifically looking for a reliable specialist for a long-term cooperation.

Application questions

Please answer the following questions:

1. Describe one dataset audit you completed for a computer-vision project.
2. How would you identify duplicate and near-duplicate frames in a long drive-by recording?
3. How would you prevent data leakage between training and test data?
4. What would you check before using LiDAR data together with RGB images?
5. How would you determine whether existing model output is suitable for pre-annotation?
6. How many hours do you estimate for this audit?
7. What fixed price do you propose?
8. When can you start?
9. What exactly will you deliver at the end of the audit?

Project Overview
We are a research team preparing an IEEE conference manuscript investigating the effects of synthetic skin-tone augmentation on dermatology image classification models. Our work evaluates whether instruction-based image editing can improve model generalization across diverse skin tones while preserving lesion morphology.

The study includes:
Synthetic-replacement ablations (0%, 25%, 50%, 75%, and 100% synthetic data)
External validation on independent dermatology datasets
Skin-tone subgroup analyses across Fitzpatrick skin types
Evaluation of model generalization and synthetic image validity

What We Need
We are seeking PhD-level reviewers with experience in one or more of the following areas:
Artificial Intelligence / Machine Learning
Medical Image Analysis
Computer Vision
Dermatology
Explainable AI or Healthcare AI
Reviewers will provide feedback on:
Methodology and experimental design
Statistical analyses and evaluation metrics
Clarity and organization of the manuscript
Interpretation of results and limitations
Suggestions for improving scientific rigor and presentation for academic venue

Preferred Qualifications
Currently pursuing or holding a PhD in a relevant field
Prior experience conducting or reviewing academic research
Familiarity with machine learning and/or medical imaging is preferred
Prior experience serving as a reviewer for IEEE-affiliated conferences or journals is highly preferred

Expected Commitment
Review one IEEE conference-length manuscript (~4-6 pages)

Provide written comments and suggestions for improvement

We welcome both major and minor critiques. Our goal is to strengthen the scientific quality, rigor, and clarity of the manuscript prior to submission and receive realistic feedback similar to peer review for an academic venue to improve our manuscript's quality.

We are seeking a skilled machine learning engineer to develop and implement computer vision models using TensorFlow.js. The ideal candidate will have experience in image classification and be able to work independently to deliver high-quality results. This is a part-time role with a project scale of small, expected to last less than a month.

I'm the founder of a humanoid robotics AI company. I need a robotics/RL engineer who can work alongside me on real problems while explaining their thinking and approach so I build deep understanding over time.

This could lean more toward hands-on engineering or more toward structured teaching depending on your strengths. Ideally both.

Scope:

• Reinforcement learning for motor control (PPO, SAC, reward shaping, curriculum learning)
• Sim-to-real transfer pipelines for legged/humanoid robots
• Physics simulation environments (MuJoCo, Isaac Sim, Isaac Lab)
• Locomotion policy training and deployment
• Control systems fundamentals (PID, impedance control, EtherCAT)
• URDF/MJCF model creation and tuning
• Real-time inference at the edge

About me:

Strong software engineering background. I work daily with a robotics team and need to go from high-level AI/ML understanding to being able to evaluate technical decisions, read papers, and contribute meaningfully to architecture discussions.

Format:

• Video calls every 1-3 days (1-2 hours per session)
• Mix of collaborative engineering work and structured learning
• Async Q&A between sessions
• Duration: 2-3 months initially, ongoing if it works well

Ideal candidate:

• Has shipped sim-to-real on real legged or humanoid hardware
• Professional experience with MuJoCo and/or Isaac Sim
• Strong communicator who can explain complex concepts clearly
• Comfortable with a collaborative working style, not just isolated task execution

Budget: Open to hourly or weekly retainer. Share your rate and availability and we can arrange a short intro call to see if we are a fit.

We are seeking a talented AI and Computer Vision specialist to join our team for an exciting project. The ideal candidate will have experience in developing and implementing cutting-edge AI algorithms and computer vision applications. You will work collaboratively to enhance our existing systems and create new features that leverage image processing and machine learning techniques. If you are passionate about technology and have a track record of delivering impactful solutions, we would love to hear from you!

I need an intermediate AI Engineer with grip on Machine learing and computer vision, having knowledge about docker and aws will be a plus, immediately required.

We are seeking a skilled freelancer to develop a computer vision system for NBA2K autogreen features. The system should automatically green shots for the user with meter detection and without a meter detection. i want the computer vision to be called “wunna vision”i also have access to someone else vision channel and there doing awsome things with something called omni. i have no idea what it is but if you can explain it to me and somehow incorporate that into mine ill add an additional 50$. this should work on ps and xbox Additionally, the project involves creating a UFC area for all ufc things like “auto deny transitions” “auto block” and a feature called “wunna engine” where when enabled it fights for you and there’s a slider that controller his aggressiveness. The ideal candidate will have experience in computer vision and scripting, with a strong understanding of image processing and automation. This is a part-time project with a short duration, expected to be completed in less than a month.

Hello Team,

We are planning to build a custom enterprise platform for my company and require an end-to-end software development cost estimation (including architectural backend engineering and frontend dashboard layout) for the following core modules:

Custom NLU Engine (Local Context): AI agent orchestration optimized for colloquial language (handling mixed Romanized Urdu dialects and phonetic spellings) with robust runtime typo-handling, multi-turn state persistence, and dynamic multi-model API fallback logic.

Vision Surveillance Integration (CV Guard): Background webcam frame processing using local Computer Vision models for real-time facial tracking, continuous desk-presence monitoring ("Admin left desk", "Unknown person detected"), and generating immediate database log event triggers.

Enterprise Security & Granular RBAC: Advanced multi-tenant permission controls separating Admin and Staff access. Features must include a highly secure, flexible 4-6 digit PIN session locking mechanism, and cascading chat-sharing toggles (with individual switches for resharing restrictions, lock bypass, and delete protections).
Task Scheduling Engine: Background worker/cron systems capable of reading user text entries inside active note logs and scheduling custom system actions, alerts, or push notifications based on custom text timestamps (e.g., triggering an automated alert after a specified time frame).


Please provide us with a detailed ballpark pricing and timeline estimate for engineering this entire secure infrastructure from scratch.

Regards.

Skills: Java, Python, Mobile App Development, MySQL, Software Development, Backend Development, Computer Vision, AI Chatbot Development, AI Development, AI Agents

I buy distressed residential properties in the midwest USA and want to automate the process of finding new opportunities.

The goal is to build a system that continuously monitors new residential listings, uses AI to analyze listing photos and descriptions, identifies properties that appear distressed, and immediately notifies me with the listing details.

This is intended to become the foundation of a much larger AI acquisitions platform, so I'm looking for someone who values clean architecture and scalable design.

What the System Should Do

Continuously monitor newly listed residential properties.
Analyze listing photos using an AI vision model to determine whether a property appears distressed.
Analyze listing descriptions for phrases and language commonly associated with distressed properties (such as "investor special," "fixer upper," "TLC," "cash only," etc.).
Flag a property if either the photos or the description indicate distress.
Send an immediate notification (SMS or email) that includes: property address, listing price, listing agent name, agent phone number, agent email, brokerage, link to the listing, and a brief explanation of why the AI flagged the property.

The target is to receive notifications within approximately five minutes of a qualifying listing becoming available.

Preferred Technology

I'm open to recommendations, but I expect the solution to involve technologies such as Python, OpenAI or Google Gemini vision models, APIs for listing data (MLS, RESO, IDX, or another reliable and compliant source), Docker, Git, Twilio or another SMS provider, and PostgreSQL or another appropriate database. The most important factor is building a reliable, maintainable system, not using a specific technology.

I do not require MLS access. I'm flexible on the listing data source and open to Zillow, Realtor.com, RapidAPI-based listing feeds, or any other approach the developer recommends, as long as it's reliable and the developer is upfront about any terms-of-service or data-reliability tradeoffs of the chosen source.

What I'm Looking For

Experience with Python development, AI integrations, computer vision, API development and integrations, automation, cloud deployment, and building production-ready applications. Experience with real estate data, MLS/IDX/RESO feeds, or property technology is a significant plus.

Deliverables

Clean, well-organized source code; documentation for setup and deployment; configuration files; error handling and logging; a system that can run continuously with minimal maintenance; and instructions for future expansion.

Timeline

I'd like to have a working Version 1 as quickly as possible, ideally within the next few weeks.

When You Apply

Please include examples of AI or automation projects you've built, any experience with computer vision or image analysis, any experience with real estate APIs or MLS/IDX/RESO data, your recommended technical approach for this project, your estimated timeline and budget, and any questions or concerns you have about the project.

I'm looking for someone interested in building a long-term relationship. If Version 1 is successful, I have additional AI automation projects planned and would prefer to continue working with the same developer.

Definition of Done (Critical Requirement)

This project is not considered complete when code is written or when a prototype is demonstrated. The final deliverable must be a fully functional, deployed Version 1 system that is actively running and automatically performing its intended purpose. When the project is delivered, I expect to be able to immediately begin receiving qualified property alerts without needing to complete additional technical setup, coding, configuration, integrations, or development work.

The completed system must: be deployed and running continuously (24/7); automatically monitor the agreed-upon listing data source; automatically analyze new listings using AI; automatically identify properties matching the distress criteria; automatically send notifications to me without manual intervention; continue operating after delivery without requiring the developer to manually run scripts or restart processes; include all required API connections, accounts, credentials, and integrations necessary for operation; and include documentation explaining how the system operates and how to maintain it.

A successful delivery means: "I wake up tomorrow, a distressed property is listed that matches my criteria, and the system automatically sends me the alert." That is the acceptance criteria.

Deployment Responsibility

The developer is responsible for setting up the hosting environment, configuring APIs, connecting all required services, setting up notifications, testing the complete workflow from listing detection to notification delivery, and confirming that the system runs automatically without my involvement.

I am not looking for a tutorial, a code repository, or a partially completed application. I am looking for a working AI-powered monitoring system.

• Which listing data source would you use — MLS/RESO, Zillow, Realtor.com, or a RapidAPI-style feed — and why?
• Where will this be hosted, and what happens if the server crashes or the API changes?
• What's your estimated monthly cost to run this (hosting, AI API calls, SMS, data feed), separate from your one-time fee?
• What could realistically cause this project to fail or run over budget, and what would you do about it?
• How would you reduce false positives — a listing that says 'TLC' but isn't actually distressed, or a renovated house with an old, cluttered-looking photo?

I have mapped out an end-to-end system that detects and segments more than eighty common food classes with YOLOv8-seg and returns calorie estimates in under a second. The architecture is already sketched: a Django REST API backed by SQLite3, a responsive HTML / CSS / JavaScript front end, and automated deployment to Render through CI/CD.

Where I now need support is turning this blueprint into a polished, user-facing product. Users must be able to:

• upload a meal photo and receive an on-screen breakdown of detected items, servings, macros and total calories
• open a nutrition panel for any item to drill into vitamins, allergens and other details
• save each analysis so their meal history accumulates automatically, ready for basic charts and summaries
• schedule a consultation slot and request an AI-assisted personalised meal plan generated from their history

A straightforward email-and-password flow will gate all of this. The computer-vision core is trained and validated; your focus will be clean API endpoints, state management on the front end, and wiring the consultation / meal-plan logic.

Acceptance criteria
1. Endpoints accept JPEG/PNG ≤8 MB, return JSON with classes, masks, weight estimates and calories in ≤1 s on Render’s standard instances.
2. Front-end renders masks as overlays, displays nutrition facts and writes meal logs without page reloads.
3. Meal history, consultation bookings and generated plans persist in SQLite3 and are exposed through paginated REST routes.
4. CI/CD pipeline deploys a tagged commit to Render and passes automated tests for inference speed (<1 s), data integrity and authentication.

Tech stack is fixed to Django 4, YOLOv8, SQLite3, Tailwind (optional), vanilla JS and GitHub Actions, but I’m open to lightweight suggestions that keep deployment simple on Render. Deliver workable, documented code and a short video demo, and we’re done.

Skills: JavaScript, CSS, Django, HTML, SQLite, Git, Computer Vision, Deep Learning, API Development, REST API

We're looking for contributors from Brazil, Mexico, Colombia, and Argentina to participate in a paid smartphone-based First-Person View (POV) Video Collection Project.

You'll record natural everyday activities using an approved smartphone mounted on a head strap. The recordings help improve advanced AI and computer vision systems.

Examples of Activities
Cleaning rooms
Laundry
Cooking
Kitchen cleanup
Organizing household items
Washing dishes
Gardening
Grocery organization
Pet care
Walking
Other natural daily activities
Requirements

✅ Must currently reside in:

Brazil
Mexico
Colombia
Argentina

✅ Approved smartphone:

iPhone 11–17 Series (16e & 17e not supported)
Google Pixel 6–9 Series
Samsung Galaxy S21, S22, S23 & S24 Series

✅ Head-mounted phone holder

Recording Requirements
Record directly in the Minute Data App.
Minimum: 2 minutes per recording.
Maximum: 30 minutes per recording.
Multiple recordings can be completed daily.
Compensation

💰 USD $4/hour for the first 50 approved production hours

💰 USD $5/hour after successfully completing 50 approved production hours

• Which country are you currently located in?
• What smartphone model do you use?
• Do you own a head-mounted phone holder?
• Can you start this week?

We're looking for contributors based in the United States to participate in a paid smartphone-based First-Person View (POV) Video Collection Project.

You'll record natural daily activities using an approved smartphone mounted on a head strap. The collected videos help train next-generation AI and computer vision systems.

Examples of Activities
Cleaning rooms
Organizing household items
Washing dishes
Kitchen cleanup
Laundry
Grocery organization
Gardening
Pet care
Walking
Other everyday household activities
Requirements

✅ Must currently reside in the United States

✅ Approved smartphone:

iPhone 11–17 Series (16e & 17e not supported)
Google Pixel 6–9 Series
Samsung Galaxy S21, S22, S23 & S24 Series

✅ Head-mounted phone holder

Recording Requirements
Record directly in the Minute Data App.
Minimum: 2 minutes per recording.
Maximum: 30 minutes per recording.
Multiple recordings can be completed daily.
Compensation

💰 USD $8 per approved recording hour

• Which state are you located in?
• What smartphone model do you use?
• Do you own a head-mounted phone holder/Head Strap ?
• Can you start this week?

We are looking for a 3d animator to create content for our company Silimind. We have a ton of video ideas we'd like to do and will hope to find a long term contract with a freelancer. First with a deadline of Monday, so working over the weekend this one time is needed!

We need a video like this for our software product SiliVision: https://youtu.be/ab2FqTbetFw?is=7LzAVZXEO7vkozzg

The script is:
Grabber Intro Clip and/or intro slide
SiliVision platform rotating: Multi-Modal AI Computer Vision System (backwards compatible)
Showcase Orin rotating with analytics breakdown (shift down to corner after done, dont fully remove)
Showcase NX rotating with analytics breakdown (shift down to corner after done, dont fully remove)
Showcase AGX rotating with analytics breakdown
Silimind Logo (Presales Start Now)


Extended Version: Add this-
Show compatible with any camera system adding a cameras and it going live
Show thousands of detections available (screen filled with falling items)
Show it's multimodality
How To Get Started
Visit SiliVision Presale Site
Showcase Questionnaire, Packages, License and marketplace quickly
Showcase Orin rotating with analytics breakdown (shift down to corner after done, dont fully remove)
Showcase NX rotating with analytics breakdown (shift down to corner after done, dont fully remove)
Showcase AGX rotating with analytics breakdown
Closing with use case generalization


I can meet with you to explain exactly what I need and what im imagining. I just dont have the time to build these right now. Please share a sample 1-3 links of your best work to see if its compatible. We want apple quality ads only! For this first ad 2 versions are needed a 10-15 second ad and an extended version. Apply, let's talk!

Hello there,

I have a new app that listed on App Store recently. I am looking for an AI expert to check & see what features can be added to the app to bring more & better look to the app.

Note: My app ONLY shows Exchange rate like Fiat, Gold, Metals & & Crypto Prices.

Thank you.

I buy distressed residential properties in the midwest USA and want to automate the process of finding new opportunities.

The goal is to build a system that continuously monitors new residential listings, uses AI to analyze listing photos and descriptions, identifies properties that appear distressed, and immediately notifies me with the listing details.

This is intended to become the foundation of a much larger AI acquisitions platform, so I'm looking for someone who values clean architecture and scalable design.

What the System Should Do

Continuously monitor newly listed residential properties.
Analyze listing photos using an AI vision model to determine whether a property appears distressed.
Analyze listing descriptions for phrases and language commonly associated with distressed properties (such as "investor special," "fixer upper," "TLC," "cash only," etc.).
Flag a property if either the photos or the description indicate distress.
Send an immediate notification (SMS or email) that includes: property address, listing price, listing agent name, agent phone number, agent email, brokerage, link to the listing, and a brief explanation of why the AI flagged the property.

The target is to receive notifications within approximately five minutes of a qualifying listing becoming available.

Preferred Technology

I'm open to recommendations, but I expect the solution to involve technologies such as Python, OpenAI or Google Gemini vision models, APIs for listing data (MLS, RESO, IDX, or another reliable and compliant source), Docker, Git, Twilio or another SMS provider, and PostgreSQL or another appropriate database. The most important factor is building a reliable, maintainable system, not using a specific technology.

I do not require MLS access. I'm flexible on the listing data source and open to Zillow, Realtor.com, RapidAPI-based listing feeds, or any other approach the developer recommends, as long as it's reliable and the developer is upfront about any terms-of-service or data-reliability tradeoffs of the chosen source.

What I'm Looking For

Experience with Python development, AI integrations, computer vision, API development and integrations, automation, cloud deployment, and building production-ready applications. Experience with real estate data, MLS/IDX/RESO feeds, or property technology is a significant plus.

Deliverables

Clean, well-organized source code; documentation for setup and deployment; configuration files; error handling and logging; a system that can run continuously with minimal maintenance; and instructions for future expansion.

Timeline

I'd like to have a working Version 1 as quickly as possible, ideally within the next few weeks.

When You Apply

Please include examples of AI or automation projects you've built, any experience with computer vision or image analysis, any experience with real estate APIs or MLS/IDX/RESO data, your recommended technical approach for this project, your estimated timeline and budget, and any questions or concerns you have about the project.

I'm looking for someone interested in building a long-term relationship. If Version 1 is successful, I have additional AI automation projects planned and would prefer to continue working with the same developer.

Definition of Done (Critical Requirement)

This project is not considered complete when code is written or when a prototype is demonstrated. The final deliverable must be a fully functional, deployed Version 1 system that is actively running and automatically performing its intended purpose. When the project is delivered, I expect to be able to immediately begin receiving qualified property alerts without needing to complete additional technical setup, coding, configuration, integrations, or development work.

The completed system must: be deployed and running continuously (24/7); automatically monitor the agreed-upon listing data source; automatically analyze new listings using AI; automatically identify properties matching the distress criteria; automatically send notifications to me without manual intervention; continue operating after delivery without requiring the developer to manually run scripts or restart processes; include all required API connections, accounts, credentials, and integrations necessary for operation; and include documentation explaining how the system operates and how to maintain it.

A successful delivery means: "I wake up tomorrow, a distressed property is listed that matches my criteria, and the system automatically sends me the alert." That is the acceptance criteria.

Deployment Responsibility

The developer is responsible for setting up the hosting environment, configuring APIs, connecting all required services, setting up notifications, testing the complete workflow from listing detection to notification delivery, and confirming that the system runs automatically without my involvement.

I am not looking for a tutorial, a code repository, or a partially completed application. I am looking for a working AI-powered monitoring system.

• Which listing data source would you use — MLS/RESO, Zillow, Realtor.com, or a RapidAPI-style feed — and why?
• Where will this be hosted, and what happens if the server crashes or the API changes?
• What's your estimated monthly cost to run this (hosting, AI API calls, SMS, data feed), separate from your one-time fee?
• What could realistically cause this project to fail or run over budget, and what would you do about it?
• How would you reduce false positives — a listing that says 'TLC' but isn't actually distressed, or a renovated house with an old, cluttered-looking photo?

Role: Full-Stack Systems & Edge AI Engineer

We are a Singapore-based startup building cutting-edge computer vision solutions for the food industry. We are seeking a hands-on, end-to-end engineer to take ownership of our pipeline—from edge-side data capture to cloud-based model training and deployment.

What You’ll Do:

Edge Pipeline Development: Architect and maintain low-latency inference pipelines using YOLO, BoT-SORT, and OpenCV to run reliably on edge hardware.

Cloud Infrastructure: Manage data flows between edge devices and GCP using rClone and Google Cloud Storage.

Data Ops: Streamline our annotation workflow using Label Studio (Dockerized) and curate high-quality datasets with FiftyOne.

Model Lifecycle: Leverage SAM 2 for advanced segmentation tasks and manage the full cycle of training, evaluation, and edge deployment.

System Integration: Develop robust, asynchronous backend services using FastAPI to connect our edge systems to the cloud.

Who You Are:

A pragmatic problem-solver who enjoys moving from prototype to production.

Experienced in navigating the trade-offs between edge hardware constraints and model performance.

Passionate about building scalable AI systems that solve real-world problems in the food industry.

I'm looking for a senior engineer to help evaluate and build a proof of concept for a browser-based application involving real-time video processing.

Due to the nature of the project, full product details will be shared after an NDA.

This project combines several advanced browser technologies and I'm looking for someone who has experience working at this level.


Experience with some or all of the following:
- Chrome Extensions (Manifest V3)
- WebRTC
- MediaStreams / MediaStreamTrack
- Webcam and microphone APIs
- Canvas video rendering
- Computer Vision
- Real-time video processing
- Browser performance optimization
- TensorFlow.js, MediaPipe, ONNX Runtime, WebGPU or similar
- JavaScript / TypeScript
- React (preferred)

The initial engagement is a paid proof of concept focused on determining technical feasibility.

Areas to evaluate include:
- Real-time webcam processing
- AI/computer vision integration
- Browser-based video rendering
- MediaStream manipulation
- Chrome Extension architecture
- Performance and latency optimization
- Local vs. cloud inference
- Overall technical feasibility

The outcome of this phase will determine whether we move forward with a full MVP.

Ideal Candidate
- Senior-level engineer
- Strong experience with WebRTC and browser media pipelines
- Experience building Chrome Extensions
- Comfortable working with experimental technologies
- Able to recommend architecture rather than simply execute requirements
- Excellent communicator

• Describe the most similar project you've built involving Chrome Extensions, WebRTC, real-time video, or AI. What was your role? Can you share examples
• Please share links to any relevant examples, demos, portfolio pieces, or public projects you've worked on.
• Can you provide 2–3 client references or examples of projects where you were hired to solve a technically complex engineering problem?

# AI Data Collection Associate (Remote)

**Company:** Adept Global Innovation Center (Adept GIC)

📍 **Location:** Remote (Work from Home)

💼 **Job Type:** Freelance | Part-Time | Flexible Hours

🎓 **Experience:** No Experience Required

🎤 **Language:** No speaking or audio recording required

## About the Opportunity

Join Adept GIC and contribute to the future of **Artificial Intelligence, Robotics, and Computer Vision**.

We're looking for participants to record **first-person (POV) videos** while performing their normal daily activities at home or work. Your recordings will help train next-generation AI systems.

This opportunity is perfect for students, homemakers, office employees, warehouse workers, retail staff, and freelancers.

## Eligible Countries

* United States
* Canada
* Mexico
* Brazil
* Colombia
* Argentina

## Device Requirements

**Apple iPhone:** 11–17 Series *(16e & 17e not supported)*

**Google Pixel:** Pixel 6–9 Series

**Samsung Galaxy:** S21–S24 Series

**Additional Requirements:**

* Head-mount phone strap
* Stable internet connection

## What You'll Do

Record natural first-person POV videos while performing activities such as:

* Cooking
* Cleaning
* Laundry
* Dishwashing
* Home organization
* Warehouse operations
* Inventory handling
* Logistics
* Retail and workplace activities

**Recording Duration:** 2–30 minutes per video

**Maximum Participation:** Up to **100 recording hours per environment**.

## Apply Now

If you have a compatible smartphone and would like to contribute to the future of AI while working on your own schedule, we'd love to hear from you!

Wix Website Build — Marketing Site, Registration & Cart, Ad Landing Pages (RITS)
Category: Web Design / Wix Development
Contract type: Hourly
Experience level: Intermediate–Expert
Platform: Wix — see platform note below
Target launch: September 1 — hard date. Fall enrollment opens that day.

About Us
We are a STEM and AI education company serving middle and high school students. We run in-person summer camps and weekly live online programs, and we hold a university partnership.

We need a sophisticated marketing site that drives camp and online-program registrations, builds credibility, opens partnership channels, and supports paid-ad campaigns.

Our students are minors — every registration is completed by a parent or guardian.
Platform & Tools
Build in Wix using native Wix tools:

- Wix Blog for the blog
- Wix Forms for forms and registration
- Wix Stores / cart and checkout for payment
- Wix Payments (Stripe) as the processor
- Zapier for automation and data routing after submission
Important constraints — please read before bidding
No Velo / custom code. This must be built with native Wix tools and configuration. We understand this limits what the registration flow can do, and we accept reasonable compromises to stay code-free.

No Wix CMS / Collections architecture. Camps, tracks, and programs are designed pages, not CMS-driven collections. Our staff edits content directly.

We use Zapier, not Velo, and we know they aren't interchangeable. Zapier handles what happens after a registration — routing data to spreadsheets, email, and our other tools. If something in this scope genuinely cannot be done natively, say so in your proposal and tell us what the workaround costs us. We'd rather adjust the requirement than discover the gap in August.

Platform: Editor or Studio? Our team works in Wix Editor. Our internal brief specifies Wix Studio for responsive control. Recommend one, given that non-technical staff maintain content afterward, and factor it into your estimate.
Brand
- Fonts: A refined serif (Playfair Display or Fraunces) for display headlines; Montserrat for all body and UI text. Two families only. This is decided — we are not looking for font exploration.
- Logo: RITS wordmark plus the arch/aqueduct icon mark. Full lockup in header and footer, compact version for mobile. Files provided.

Design Direction
Sleek, sophisticated, editorial. A design-forward university brand crossed with a modern AI company: generous whitespace, strong hierarchy, large confident headlines, and dramatic dark cobalt sections between lighter ones for rhythm. Subtle, tasteful motion — fade-and-rise on scroll, gentle hover lifts. Use the arch shape as a recurring signature accent.

Avoid a childish camp look. Avoid a generic template feel. High-quality, cool-toned photography throughout — placeholders acceptable for now.

With custom code off the table, design quality is where this project is won. We're hiring for taste as much as build skill.


Scope of Work
Part 1 — Main Site

Home — Hero with primary CTA, value pillars, program overview (Summer Camp + Online Programs), credibility strip (university partnership, logos), testimonials and outcome stats, recent blog previews, newsletter capture, closing CTA.
About Us — Required. Mission, instructors, facility, credibility — and a partnership inquiry section with its own form.
Summer Camp — Holds two camps (Middle School, High School), each with its own tracks — e.g. AI, Computer Vision, Vibe Coding & CS, Biomedical Sports Science. Dates, price, ages, schedule, what's included, faculty, gallery, FAQ, and a registration CTA.
Online Programs — Weekly live online classes, middle and high school. Tracks/cohorts, how it works, instructors, pricing, testimonials, FAQ, enroll CTA.
Partnerships — Dropdown menu with three sub-pages sharing one layout (intro, offer, who it's for, credibility, inquiry CTA): Private & Home Schools · Speaking & Presentations · University (anchored by the VCU College of Engineering relationship).
Blog — Wix Blog, client-editable. Editorial post grid, category filters, clean post template. Seed with a few placeholder posts.
Contact Us — Contact form with reason dropdown, email and phone, map, social links, response-time note.


Part 2 — Registration, Cart & Checkout
Replaces our current external Google Form.

Multi-step registration flow with a cart:

1. Select program — Summer Camp (by camp and track) or Online Program
2. Student and guardian information
3. Add additional classes as optional line items
4. Review cart
5. Pay
6. Confirmation

Requirements:

- Payment via Wix Payments (Stripe)
- Completed by a parent or guardian, not the student
- Guest checkout — no member accounts or login. Parents register and pay without creating an account.
- Multiple classes in one cart. A parent should be able to add more than one class and check out once. We must be able to add new class offerings over time without rebuilding checkout.
- Flat fee per program, regardless of enrollment date. No prorating.
- Rolling enrollment for Fall online programs — the flow stays live through the semester and handles students joining mid-semester
- Capacity limits per camp/track, with a graceful "full" state [waitlist? — TBD]
- Registration data routed via Zapier to [Google Sheets / our email tool / TBD] so our staff can work with it
- Working contact form, partnership inquiry forms, and newsletter email capture — all routed appropriately

Planned for a later phase — don't design us into a corner. We will need to capture a parent-signed waiver and media release at registration, electronically signed and retrievable. It is not in scope for the September 1 launch — we're handling waivers outside the site for now. But we intend to add it. Please don't architect the registration flow in a way that makes it painful to bolt on later, and tell us in your proposal how you'd eventually add it (marketplace app, external e-sign tool via Zapier, or otherwise).
Part 3 — Mobile-First Ad Landing Pages
Standalone pages for paid ads (Meta, Instagram, Google), separate from the main site menu, on their own URLs:

- One for Summer Camp, one for Online Programs, optionally a lead-capture/waitlist page
- Mobile-first, single goal, one primary CTA, minimal navigation, sticky mobile CTA, fast loading
- Conversion-focused sections: benefit-led hero, social proof, what's included, pricing and value, urgency, short FAQ, final CTA
- Built as an easily duplicated template for future campaigns


Standards
- Fully responsive and mobile-first
- Accessible — good contrast, alt text, labeled forms
- Fast-loading
- Clean page structure and social share images
Out of Scope
- Velo / custom code — native Wix only
- Wix CMS / Collections architecture
- Member accounts / login — guest checkout only
- Waiver / e-signature capture — later phase, not this build
- Logo design or brand identity work — assets and specs provided
- Typography selection — fonts are decided
- Style guide — not needed
- SEO strategy or optimization services — not needed (basic structural hygiene only)
- Copywriting — we provide all text
- Photography and videography — placeholders acceptable for now
- Migration off Wix
- Ongoing content updates post-launch (separate maintenance arrangement possible)
What We Provide
- Wix account access with appropriate permissions
- Zapier account access
- Logo files, color specs, font specs
- All written copy and program details
- Any existing photos
- A single point of contact for feedback and approvals, available for fast turnarounds

Note: All dates, prices, ages, names, stats, and testimonials are placeholders and will change. Keep them easy to edit.
Requirements
- Demonstrated Wix expertise — link 2–3 live Wix sites you've built, ideally with Wix Stores or multi-step forms
- Wix Stores / cart and checkout configuration experience
- Wix Forms multi-step form experience
- Wix Payments / Stripe integration experience
- Zapier experience connecting Wix to outside tools
- Strong editorial/visual design sense. Portfolio should show work that looks designed, not templated. This matters more than anything else on this list.
- Paid-ad landing page / conversion design experience is a real plus
- Genuine availability between now and September 1. Please be honest about your workload.
- Clear written communication, responsive during Eastern Time business hours
- Bonus: education, nonprofit, or youth-program clients
Engagement Details
- Hourly contract, tracked via Upwork's time tracker, with a weekly hour cap agreed before we start
- Target start: as soon as we find the right person · Hard launch: September 1
- Weekly check-in call plus async updates
- Before we begin, we want a rough hour estimate broken out by phase (design, main site build, registration + cart, Zapier automations, landing pages, QA) and a heads-up if you're trending over
- We expect to phase this. If the full scope can't land by September 1, our priority order is: Online Programs + registration and cart → Home → About Us → Summer Camp → ad landing pages → Partnerships → Blog. Tell us if you'd sequence it differently.
Deliverable
A complete, published, responsive Wix site with all pages above, a working Partnerships dropdown, a client-editable blog, a functional multi-step registration flow with a cart and Stripe payments, Zapier-routed submissions, working contact, partnership inquiry, and newsletter capture, and the standalone mobile-first ad landing pages — all consistent with the brand and design direction above.
How to Apply
Please include:

1. 2–3 relevant Wix portfolio links (live sites, not mockups)
2. Your hourly rate and honest weekly availability between now and September 1
3. A rough hour estimate by phase
4. Your recommendation: Wix Studio or Wix Editor, and why
5. How you'd build a multi-step registration with a cart using native Wix only — and anything in this scope you believe can't be done without custom code
6. How you'd eventually add parent-signed waiver capture in a later phase
7. Any red flags — especially on the timeline. We'd rather hear it now than in week three.

Start your proposal with the word "Richmond" so we know you read the whole thing.

We're reviewing on a rolling basis and moving quickly.

Are drones with lasers a possible solution for pest control? I want to have a panel discussion about this on YouTube.

Basically, to have a critical discussion of this: www.photonicinsecticides.com

Each panel member will have validated and discus a different aspect:
- eye safety
- energy requirements and cost per hectare
- cost comparison to cart based systems
- the aiming loop
- and biological reality: how does the target insect behave etc

Requirements: have a critical mind. Speak English well. Have some science degree. Be able to work with AI (Claude, codex, …) so that you can use those to validate the concepts and numbers.

Appreciated experience: computer vision expert, robotics, gaming or simulation.

The goals it to reduce the level of sceptisicm that people about about the possibility of this technology becoming real. I try this by having independent people speak free after having heard to arguments (and taking some time to have Claude etc validate the calculated numbers).

This is a bit of a special project. Hopefully I can find the right people for this.

Format: recorded zoom or google meet meeting.

Kr,
Nick

Looking for an ML engineer to develop and implement a Model for real-time blood pressure (BP) estimation using heart rate photoplethysmograph (PPG) signals which will be deployed on an MCU for non-invasive health monitoring.

Responsibilities:
• Design and implement a robust algorithm for extracting BP values from raw or pre-processed PPG data (depending on the provided format).
• Utilise machine learning or neural network techniques to achieve accurate BP estimation.
• Optimise the algorithm for real-time performance on Nordic chipsets in an embedded C environment (or provide an embedded C based implementation of the model)
• Integrate the algorithm seamlessly with existing hardware and software platforms (will be done by our engineers)
• Conduct extensive testing and validation of the algorithm's accuracy and performance under various conditions. (will be done by our engineers)
• Document the algorithm design, implementation, and testing process clearly and concisely.


Ideal candidate:
• Strong foundation in Python, C, C++
• Solid understanding of signal processing and machine learning/neural network techniques.
• Experience with real-time data processing and embedded system optimization.
• Familiarity with physiological signals, particularly PPG and blood pressure dynamics, is a plus.
• Excellent analytical and problem-solving skills with a meticulous approach to detail.
• Strong communication and documentation skills.


References:
(link removed)

Join our team to build 3 AI agents for a Lego assembly machine. The project involves taking a Lego build, identifying the pieces, and creating a life-like assembly process. We are looking for someone with expertise in AI and automation, particularly in machine learning and computer vision. Experience in robotics and mechanical engineering is a plus.

Algorithm Developer – Automatic Cadastral Map Vectorization (GIS / Computer Vision / AI)

We are looking for an experienced Algorithm Developer to build an automated pipeline for vectorizing scanned cadastral maps (PDFs/images) into GIS-ready vector data.

The objective is not manual digitization, but the development of a robust algorithm capable of extracting parcel boundaries and converting them into clean vector geometries with minimal human intervention.

Responsibilities

Develop an automated cadastral map vectorization pipeline.
Detect and extract parcel boundaries from scanned PDF/raster maps.
Convert raster boundaries into topologically correct vector polygons.
Handle line intersections, junctions, gaps, and noisy scans.
Implement geometry refinement, snapping, and topology correction.
Export results to GIS formats (Shapefile, GeoPackage, GeoJSON, DXF).

Hi, we're looking for a talented software developer to join Marcr, an early-stage sports technology startup based in Australia. For the right person, this role has the opportunity to eventually develop into a permanent Technical Co-Founder position, with the chance to help shape the technology, product and future direction of the company.

Marcr is being built to turn everyday football training into measurable performance data. The vision is to use computer vision, video analysis and intelligent movement tracking to analyse athletes completing real-world training drills and provide practical feedback on speed, agility, movement and performance.

We have already begun validating the concept with athletes and football clubs and are now looking for the right technical person to help build the technology and shape the product from the ground up.

We are looking for someone who sees the opportunity in what we are building, has a genuine passion for improving technology in football training and to potentially become part of the company as a co-founder, with long-term involvement in Marcr.

The ideal person would have experience or a strong interest in areas such as:

• Mobile app development
• Computer vision and video analysis
• AI and machine learning
• Human movement or pose tracking
• iOS and Android development
• Building and launching early-stage products

You do not need to be an expert in every area. More important is having strong technical ability, curiosity, and the drive to build something from an early stage.

You would work directly with our team to determine the technical direction of Marcr, build the first scalable version of the platform, test it with real athletes and clubs, and help shape the company as we grow.

If you are a developer who has wanted to build a product for athletes we would love to have a conversation.

Please tell me a little about your background, what you have built before, and why joining an early-stage sports technology company interests you.

I’m building a localized image set to fine-tune a computer-vision model and need clear photographs of Australian and New Zealand driver’s licences taken in varied lighting. If you hold either—or ideally both—licence types and can shoot sharp, glare-free images, you can help.

What I need from you
• A small set of high-resolution photos of the physical licence card shot in natural, artificial and mixed light.
• Each shot must show the entire card, all edges in frame, in focus, with no heavy reflections or filters.
• Files delivered as unedited JPEG or PNG, straight out of the camera or phone.

Your images will be used solely to study card layout, borders and background patterns; personal data will be masked before the images enter the training pipeline.

Acceptance
I’ll review each photo for focus, framing and lighting consistency. Once approved, the task is complete and payment is released.

If you’re ready to contribute, tell me which licence(s) you have and how many lighting setups you can comfortably provide, then I’ll discuss with you via email.

Skills: Image Processing, Data Collection, Image Analysis

Hi,

We are a PropTech startup developing IRIS, an Unreal Engine 5 application used by real estate developers to present apartments interactively in sales showrooms.

We are looking for an experienced AI Backend / Machine Learning Engineer to build the backend infrastructure for a new premium feature called IRIS AI Photo Studio.

((Important: You do NOT need Unreal Engine experience. Another developer is responsible for the Unreal client. Your responsibility is the AI backend and image generation pipeline.))

Goal:
Users create a screenshot inside IRIS.

They select presets such as:
* Time of day:
* Season:
* Weather:
* Mood:
* Lifestyle Story:

The backend sends the screenshot to an AI image editing model and returns a new marketing-quality visualization while preserving the original architecture.

Responsibilities:
* Design and build the AI backend
* Integrate image editing APIs (FLUX, OpenAI Images, Stability AI or similar)
* Build a prompt generation system
* Design provider-independent architecture
* Authentication
* Image upload/download
* Job queue
* Logging / Signing up
* Credit system
* Image storage
* API documentation

Required Skills
* Python (FastAPI) or Node.js
* REST APIs
* AI Image Generation / Image Editing
* Diffusion models
* Prompt Engineering
* Docker
* PostgreSQL / Supabase
* AWS S3 / Cloudflare R2
* Authentication
* Async backend architecture

Nice to Have
* FLUX
* Stability AI
* OpenAI Images API
* ComfyUI
* ControlNet
* Segmentation
* Image masking
* Computer Vision

What We’re Looking For:
We’re not looking for someone who only knows how to call an AI API. We’re looking for someone who has already built AI-powered image generation products or production AI pipelines.

Please include:
* Relevant portfolio
* Similar AI projects
* GitHub (if available)
* Tech stack you would recommend
* Estimated architecture for this system

Long-term collaboration is possible.

We are seeking a pose estimation engineer to develop software and camera hardware for capturing footage of people lifting weights. The project involves figuring out A) what exercise the lifter is doing B) How many reps C) how much weight is on the bar d) Is their pose correct from a form perspective.

I'm early on this project but I have some gyms interested. I need someone to consult with so that I make the correct early decisions on which cameras to buy and where to put them. I also need someone to help me make sure the training data we collect is generalizable and effective.

I imagine this as a weekly call, which later becomes a full or part time ML job as the company moves forward. This is my second startup... I sold the first to private equity, so I know what I'm doing, but neither me nor my cofounder is a specialist in computer vision, although my cofounder has some experience. I need a true expert to help us avoid pitfalls.

• Please explain one project where you estimated poses so that I know you are truly expert in this category of computer vision
• Do you know about camera hardware? Will you be able to help me choose the cameras to collect the initial training data?

3D LiDAR Point Cloud Bounding Box Annotation (Object Detection)

Project Description:
Overview:

We are looking for experienced data annotators to perform high-quality 3D LiDAR bounding box labeling (cuboid annotation) on point cloud datasets for autonomous driving/robotics perception models.

Scope of Work:

Annotate Objects: Draw tight, precise 3D bounding boxes (cuboids) around target objects in point cloud scenes.

Classes to Label: Vehicles (cars, trucks, buses), Pedestrians, Cyclists, and Miscellaneous obstacles.

Orientation & Heading: Correctly align the orientation vector (yaw/heading direction) of each bounding box.

Temporal Tracking (Interpolation): Track objects across sequential frames, ensuring consistent object IDs and smooth transitions.

Strict Quality Control: Adhere to strict labeling guidelines regarding box tightness, minimum point count thresholds, and ground-level alignment.

Required Skills & Experience:

Proven experience with 3D Point Cloud / LiDAR annotation (not just 2D image labeling).

Familiarity with standard annotation platforms (e.g., CVAT, Supervisely, Scale AI, LabelCloud, BAT, or similar tools).

Strong spatial awareness and understanding of $X, Y, Z$ coordinate systems and pitch/yaw/roll rotations.

High attention to detail—accuracy and consistency are critical for this project.

Project Details:
Dataset Format: .pcd or .bin files.
Volume: Initial batch of [Insert number, e.g., 500] frames. Successful completion will lead to long-term, ongoing work.
Tool: We will provide access to the labeling platform [Or specify: "Please specify which LiDAR annotation tools you are proficient in"].

How to Apply:
To be considered, please reply with:

A brief summary of your 3D LiDAR annotation experience.
A screenshot or link to past 3D annotation work (if available).
Your availability (hours per week) and your estimated rate per 100 frames.

Note: Shortlisted candidates will be asked to complete a quick, 3-frame qualification test to demonstrate their labeling accuracy.

Skills: 3D Rendering, 3D Modelling, 3D Animation, OpenGL, Computer Vision, 3D Visualization, Object Detection, AI Development

Summary
Description:
We’re looking for an experienced AI Engineer or "Next-Gen" Full-Stack Developer to lead a 1-hour guided conversation with a 15-year-old homeschooling student.

The session should provide an engaging, hands-on introduction to how modern AI is transforming software development — shifting from traditional programming to prompt-driven development and autonomous agentic workflows.

The goal is to demystify how apps are built today using AI, moving past basic chatbot interactions to show the real-world logic of how software is designed, tested, and deployed using advanced AI tools.

Suggested Topics to Cover:

- Advanced Prompt Engineering for Coding: How to shift from simple questions to structuring high-level logical instructions that guide an AI to generate clean, functional code.
- The Power of Agentic Workflows: What an AI "agent" is, how it differs from a standard LLM, and how agentic systems autonomously plan tasks, use development tools, read error logs, and debug their own code.
- Next-Gen Development Tools: A brief look or demonstration of how cutting-edge tools (e.g., Cursor, v0.dev, Bolt.new, Replit Agent) allow developers to build entire web or mobile applications using natural language.
- The Evolving Role of a Developer: How the day-to-day job of a software engineer is shifting from syntax writing to system architecture, logic design, and problem-solving.

(You’re welcome to adapt or expand these topics based on your background.)

Requirements:

* Professional experience as an AI Engineer, LLM Application Developer, or Full-Stack Developer utilizing advanced AI coding workflows.
* Deep familiarity with agentic frameworks and modern AI-assisted development tools.
* Strong communication skills, with the ability to explain abstract logic and software concepts to a younger audience.
* Comfortable leading an interactive, conversational session (ideally featuring a short, live visual demonstration of building a simple feature or app using AI).

Logistics:

* Date: Friday, June 5 or Monday, June 8
* Time: 9am Pacific Time, US
* Duration: 1 hour
* Platform: Google Meet
* Budget: $40

Application Instructions:
If you’re excited about sharing your knowledge and experiences with an eager young learner, please apply with:

* A short summary of your professional background and the AI development tools you use daily.
* Your availability for the session.
* A brief idea of a simple, visual concept or mini-app workflow you could demonstrate during the session.

About Senpai & Sakugami

Senpai (https://senpaiapp.com) started as a dating app for anime fans — helping people break the ice and connect over their shared passion for anime. We're now pivoting into a social UGC platform in the spirit of TikTok, built to empower anime fans to share their passion with like-minded creators and communities.

Our new platform, Sakugami, takes this further — empowering anime fans to become creators themselves, using generative AI to produce long-form content alongside traditional animators and artists.

The Role

We're looking for an experienced Senior AI Engineer to help power the GenAI backbone of both Senpai and Sakugami.

In the near term, you'll build the pipelines that let us integrate existing models like **SeeDance** and **Midjourney** into our product.

Longer term, we're keen on developing our own VAM (video/audio model) built specifically for anime — and we need someone who can help take that from idea to reality.

This role spans both integration work and original model development, so you'll need to be comfortable operating across the full stack of applied AI engineering — from wiring up third-party APIs to training and fine-tuning models of your own.

### What You'll Do

- Build and maintain pipelines integrating existing AI models (e.g. SeeDance, Midjourney) into our product

- Help design and develop our own anime-specific VAM (video/audio model)**

- Train, fine-tune, and evaluate video and/or audio generation models

- Curate and manage training datasets, with particular attention to anime-style visual and motion data

- Optimize model performance and inference cost for production use

- Collaborate with product and design to translate creative goals into technical model requirements

### What You'll Need

- 2-3+ years of experience building AI models, particularly in video or audio generation
- Deep experience with Python
- Strong experience with deep learning frameworks (PyTorch, TensorFlow)
- Experience with modern generative architectures relevant to video/audio (e.g. diffusion models, transformers, VAEs)
- Experience with **dataset curation and preprocessing** for model training
- Comfort working with **GPU-accelerated training** and optimizing for training/inference efficiency
- Familiarity with integrating third-party AI APIs into production pipelines

Ground-floor opportunity to build a proprietary anime-focused generative model from scratch, not just integrate someone else's

Makeshift Software Inc. is the developer of Senpai and Sakugami.