Optimize Deep Learning Projects on Upwork

Looking to develop Ai bot to automate office work and processes

Description:

About ONE EDU:
We are building ONE EDU — an adaptive AI-powered conversational mentor that helps kids build real-world skills: emotional intelligence, entrepreneurship, financial literacy, problem-solving, and business thinking. Instead of traditional lessons, children interact daily with a voice-based mentor who guides them through real-world life scenarios, decisions, and challenges.

This is NOT a basic chatbot or tutoring app — this is a system designed to help kids think critically, solve problems, and build confidence in decision-making.

The Role:
We are hiring an AI Prompt Engineer to design and build the conversational scaffolds behind our MVP. You’ll help architect the mentor's personality, design multi-turn decision trees, and build real-world micro-scenarios for kids to interact with.

This is not a one-shot prompt job. This is structured system prompt design for multi-turn adaptive conversations.

Responsibilities:

Write system prompts for a daily-use child mentor system (GPT-4o based)

Build micro-challenges and decision trees that simulate real-world situations (ex: business decisions, emotional conflicts, negotiation, financial choices)

Design Socratic dialogue flows — mentor asks questions, adapts based on child’s choices

Build edge-case handling and safe fallback responses

Calibrate tone, safety, and age-appropriate language for children 8–13 years old

Collaborate directly with founder (me) to refine educational approach

Ideal Skillset:

GPT-4o API experience required (not just ChatGPT usage)

Experience designing multi-turn, few-shot or chain-of-thought prompt chains

Ability to create conversational tone and adaptive flows (not just fact delivery)

Experience with educational or coaching-focused prompt design is a major plus

Familiarity with safety guardrails, ethical prompt handling, edge-case control

Strong written English fluency and natural language control

Our Stack (for context):

GPT-4o (OpenAI API)

ElevenLabs (TTS)

Whisper (STT)

Supabase (backend)

Vercel (frontend hosting)

LiveKit (voice pipeline)

You do not need to code for this role — but experience collaborating with developers is helpful.

Engagement Type:

Start with Paid Test Project to evaluate fit (small, focused deliverable)

If successful → full contract to complete Sprint 1 prompt library build

Budget:

Hourly — $50–100/hr range preferred

Paid test — $250–500 fixed price for first deliverable

How To Apply:

Provide 2–3 samples of prior prompt engineering work (multi-turn flows highly preferred)

Briefly describe your experience designing prompts for coaching, education, or adaptive conversations

Confirm your GPT-4o API experience level

Confirm you’re comfortable designing age-appropriate conversational systems for kids

Looking forward to seeing your work!

• Please describe your experience designing multi-turn, adaptive conversational prompts for GPT-4o (or other LLMs). Include a brief example if possible.
• Do you have experience designing age-appropriate conversational flows for children or educational use cases? Please explain briefly.
• Confirm your direct API-level experience with GPT-4o, not just using ChatGPT. Please list any platforms, SDKs, or APIs you’ve worked with.

We are seeking an experienced quant developer to optimize our existing reinforcement learning trading agent. The ideal candidate will have a strong background in quantitative finance, machine learning, and programming. You will work on improving the agent's performance through algorithmic enhancements and data-driven strategies. If you have a passion for trading systems and a proven track record in optimizing complex models, we would love to hear from you.

We are looking for an AI engineer for a fashion related app who can research relevant models depending on the requirement and check the feasibility of implementing them.

We would like some one to investigate virtual tryon, image correction.

Job Title: AI Agent Development Specialist - Healthcare Professional Recruitment Automation
About Our Company
We are a leading primary market research company specializing in global healthcare studies and pharmaceutical research. Our organization conducts comprehensive market research across multiple therapeutic areas, working with healthcare professionals worldwide to gather critical insights that shape medical innovations and healthcare policies.

Project Overview
We seek an experienced AI automation specialist to develop a sophisticated artificial intelligence agent that will revolutionize our healthcare professional recruitment and engagement processes. This AI-powered system must handle the complete lifecycle of HCP (Healthcare Professional) recruitment, from initial data collection through survey completion and interview scheduling, while maintaining the highest standards of healthcare compliance and data security.

Core Automation Requirements
Phase 1: Intelligent Data Collection & HCP Database Creation
Automated Data Aggregation:

Develop AI systems to automatically collect healthcare professional profiles from multiple sources including medical registries, professional licensing boards, hospital directories, and professional networking platforms

Implement intelligent data parsing to extract key information including medical specialties, practice locations, years of experience, board certifications, and contact information

Create automated validation systems to verify license status, practice credentials, and professional standing

Build deduplication algorithms to identify and merge duplicate profiles across different data sources

Establish real-time data enrichment processes to continuously update professional information and contact details

AI-Powered Profile Enhancement:

Deploy natural language processing to analyze professional biographies, published research, and practice descriptions

Automatically categorize healthcare professionals by therapeutic expertise, research interests, and demographic criteria

Generate comprehensive professional profiles with standardized data formats for consistent targeting

Implement scoring algorithms to rank professionals based on relevance to specific research studies

Create automated tagging systems for therapeutic areas, geographic regions, and professional experience levels

Phase 2: Multichannel Outreach Campaign Automation
Personalized Communication Generation:

Develop AI systems to create highly personalized recruitment messages tailored to individual professional backgrounds

Generate dynamic email templates that adapt content based on medical specialty, practice type, and geographic location

Create intelligent subject line optimization to maximize email open rates across different professional segments

Implement A/B testing frameworks for continuous optimization of message effectiveness and engagement rates

Build automated translation capabilities for international recruitment campaigns across multiple languages

Multi-Platform Outreach Coordination:

Establish automated email campaign systems with intelligent scheduling based on professional time zones and preferences

Implement SMS outreach capabilities with compliance-aware messaging for mobile engagement

Create professional network outreach automation for platforms frequented by healthcare professionals

Develop coordinated follow-up sequences across all communication channels with intelligent timing optimization

Build automated response tracking and engagement scoring across all outreach platforms

Phase 3: Intelligent Response Processing & Qualification
AI-Powered Response Analysis:

Implement natural language processing to automatically categorize and prioritize incoming responses from healthcare professionals

Develop sentiment analysis capabilities to identify enthusiastic participants versus reluctant or evasive responses

Create automated qualification scoring based on response content, professional background, and engagement level

Build intelligent conversation handling for common questions about study requirements, compensation, and time commitments

Establish automated screening processes to verify professional credentials and study eligibility criteria

Dynamic Conversation Management:

Develop conversational AI capabilities to handle initial inquiries and provide study information automatically

Create intelligent FAQ systems that can answer common questions about research participation requirements

Implement automated consent process management with digital signature collection and storage

Build automated reminder systems for incomplete applications or missing documentation

Establish escalation triggers for complex inquiries requiring human intervention

Phase 4: Automated Survey Distribution & Management
Survey Deployment Automation:

Create automated survey link generation and distribution systems triggered by participant confirmation

Implement personalized survey experiences based on professional specialty and demographic information

Develop automated reminder systems for incomplete surveys with intelligent timing and frequency optimization

Build progress tracking capabilities to monitor survey completion rates and identify potential issues

Establish automated data quality checks to flag incomplete or inconsistent survey responses

Survey Optimization & Analytics:

Implement real-time survey performance monitoring with automated adjustments for low completion rates

Create adaptive survey branching based on professional background and previous responses

Develop automated survey length optimization to maximize completion while gathering required data

Build intelligent question sequencing to maintain participant engagement throughout the survey process

Establish automated data validation and cleaning processes for survey response quality assurance

Phase 5: Interview Scheduling & Calendar Management
Automated Scheduling Coordination:

Develop intelligent calendar integration systems that sync with multiple calendar platforms and time zone management

Create automated availability matching between healthcare professionals and research team members

Implement smart scheduling algorithms that optimize interview timing based on professional preferences and availability patterns

Build automated confirmation and reminder systems for scheduled interviews with multiple notification touchpoints

Establish automated rescheduling capabilities for conflicts with intelligent alternative time slot suggestions

Interview Process Optimization:

Create automated pre-interview briefing systems that send relevant study materials and preparation instructions

Develop intelligent interview format selection based on participant preferences and technical capabilities

Implement automated technical setup verification for virtual interviews including platform testing and troubleshooting

Build automated post-interview follow-up systems for thank you notes, incentive distribution, and additional study opportunities

Establish automated no-show management with intelligent rescheduling and replacement participant sourcing

Phase 6: Intelligent Escalation & Human Handoff
Smart Escalation Triggers:

Develop AI systems to identify complex cases requiring human intervention based on response patterns and engagement indicators

Create automated flagging systems for potential compliance issues, participant concerns, or unusual response patterns

Implement intelligent priority scoring for escalated cases based on urgency, participant value, and study requirements

Build automated case documentation systems that provide complete interaction history for human agents

Establish automated notification systems to alert appropriate team members of escalated cases with relevant context

Human Agent Integration:

Create seamless handoff processes that provide human agents with complete participant interaction history and context

Develop intelligent case routing based on team member expertise and availability

Implement automated workflow pause capabilities when human intervention is required

Build collaborative tools that allow human agents to provide feedback for AI system improvement

Establish automated workflow resumption capabilities once human intervention is complete

Compliance & Security Requirements
Healthcare Data Protection:
Implement comprehensive HIPAA compliance measures for all data collection, storage, and transmission processes

Establish GDPR compliance frameworks for international healthcare professional data handling

Create automated consent management systems with opt-out capabilities and preference tracking

Develop encrypted data storage and transmission protocols for all sensitive professional information

Build comprehensive audit trail systems for regulatory compliance and data governance requirements

Professional Ethics & Standards:
Implement automated compliance checking for professional communication standards and medical ethics guidelines

Create automated screening for conflict of interest detection and disclosure management

Develop intelligent content filtering to ensure appropriate professional communication tone and content

Establish automated documentation for all participant interactions and consent processes

Build automated reporting systems for regulatory compliance and internal quality assurance

Technical Specifications
AI & Machine Learning Requirements:
Advanced natural language processing capabilities for content generation and response analysis

Machine learning algorithms for continuous optimization of outreach effectiveness and participant engagement

Predictive analytics for participant likelihood scoring and campaign performance optimization

Computer vision capabilities for document processing and credential verification

Deep learning models for intelligent conversation handling and automated decision-making

Integration Capabilities:
Robust API development for integration with existing CRM and research management systems

Database connectivity for healthcare professional registries and internal participant databases

Email platform integration for automated campaign management and deliverability optimization

Calendar system integration for seamless scheduling coordination across multiple platforms

Communication platform integration for SMS, professional networks, and messaging applications

Performance & Scalability:
System architecture capable of handling thousands of simultaneous healthcare professional interactions

Automated load balancing and resource allocation for peak usage periods

Real-time performance monitoring with automated optimization and error handling

Scalable infrastructure that can adapt to varying research study volumes and requirements

Automated backup and disaster recovery systems for data protection and business continuity

Deliverables & Success Metrics
Core Deliverables:
Complete AI agent system with all automation phases fully implemented and tested

Comprehensive documentation including system architecture, user guides, and maintenance procedures

Training materials and knowledge transfer sessions for internal team members

Testing protocols and quality assurance frameworks for ongoing system reliability

Integration with existing research management systems and workflow optimization

Success Metrics:
80% reduction in manual recruitment effort compared to current processes

300% increase in healthcare professional outreach capacity without additional human resources

60% improvement in survey completion rates through intelligent engagement optimization

90% reduction in scheduling coordination time with automated calendar management

95% compliance rate with healthcare data protection and professional ethics standards

Proposal Requirements
Technical Approach:
Detailed system architecture proposal with recommended technology stack and automation platform selection

Comprehensive implementation timeline with specific milestones and deliverable schedules

Technical specifications for AI capabilities, integration requirements, and scalability considerations

Compliance strategy for healthcare data protection and professional ethics requirements

Quality assurance and testing protocols for system reliability and performance optimization

Experience Documentation:
Portfolio of similar AI automation projects in healthcare or regulated industries

Case studies demonstrating experience with large-scale professional recruitment or engagement systems

References from previous clients with comparable automation requirements and compliance needs

Demonstration of expertise in healthcare data handling and regulatory compliance implementation

Evidence of successful AI system development with natural language processing and machine learning capabilities

Platform Recommendation:
Recommended automation platform with justification for healthcare professional recruitment applications

Comparison analysis of potential platforms with pros and cons for this specific use case

Integration capabilities assessment for existing healthcare research management systems

Scalability analysis and future expansion possibilities for additional therapeutic areas and geographic regions

Cost-benefit analysis of recommended platform including licensing, development, and maintenance considerations

Selection Criteria
Proposals will be evaluated based on:

Technical Expertise (35%): Demonstrated AI development capabilities and healthcare industry experience

Compliance Knowledge (25%): Understanding of healthcare data protection and professional ethics requirements

Automation Innovation (20%): Creative approaches to healthcare professional engagement and recruitment optimization

Project Management (15%): Realistic timeline planning and milestone-based delivery approach

Cost Effectiveness (5%): Value proposition and return on investment potential

About This Opportunity
This project represents a transformative opportunity to revolutionize healthcare market research through cutting-edge AI automation technology. The successful developer will create systems that directly impact medical research efficiency and healthcare professional engagement, contributing to faster drug development and improved patient care outcomes worldwide.

Hi,
I'm looking for someone who could perform a VBM (voxel-based morphometry) analysis on my T1 weighted MRI images and give me the grey matter volume levels in specific cortical areas.

You mentioned familiarity with MATLAB, so was wondering if you might be able to do this for me and if so, how much you would charge.

Django B.

I need someone who is good in matlab , deep learning and can work on real time simulation. Collection of data and report writing is also part of this work.
Budget is already mentioned and after completion of project, the budget will be release. Please don’t ask for it again and again. Only contact who is serious. Thanks

I'm seeking a freelancer to help enhance my public speaking and delivery for my PhD presentation.

Key Focus Areas:
- Public Speaking: Improve my delivery skills.
- Confidence Building: Help me overcome anxiety.

Preferred Methods:
- Practical exercises and rehearsals.

Ideal Skills and Experience:
- Experience in coaching public speaking.
- Background in academic presentations.
- Ability to provide constructive feedback.

Looking forward to your proposals!
I am currently preparing for my PhD pre-submission presentation on the topic of deep learning techniques for eye disease classification and segmentation. I am looking for an experienced communication coach or presentation expert who can:

Listen to my existing PowerPoint presentation (approx. 2 hours of content)

Suggest improvements in content delivery, language, clarity, and slide explanation

Help me improve my spoken English and communication skills for academic presentations

Conduct at least 2–3 mock sessions with constructive feedback

Guide me in making my tone more confident and clear for a professional academic audience

I prefer someone who has experience with research/academic presentations, especially in the domain of computer science, AI, or healthcare.

Please apply with:

Your experience related to academic or research communication

Any sample feedback work (if available)

Your availability over the next 2 weeks

Skills: Research, Powerpoint, Technical Writing, Education & Tutoring, Research Writing

Summary
e are looking for a highly skilled Python developer with proven experience in building agent-based AI applications using LangChain and LangGraph. This is an exciting opportunity to work on a cutting-edge AI project involving intelligent agents, reasoning systems, and dynamic workflows.

Our goal is to build a robust AI-powered system that leverages multi-agent collaboration and knowledge graphs to perform complex tasks such as data analysis, decision-making, medical assistance (e.g., dental diagnosis), and claims processing automation.


Key Responsibilities:
Design and implement agentic workflows using LangChain and LangGraph.

Build and deploy autonomous agents capable of task planning, execution, and collaboration.

Integrate external tools, APIs, and vector databases (e.g., Pinecone, FAISS, Chroma).

Optimize agent memory, tool use, and feedback loops for improved reasoning.

Create modular, scalable, and well-documented Python code.

Summary
e are looking for a highly skilled Python developer with proven experience in building agent-based AI applications using LangChain and LangGraph. This is an exciting opportunity to work on a cutting-edge AI project involving intelligent agents, reasoning systems, and dynamic workflows.

Our goal is to build a robust AI-powered system that leverages multi-agent collaboration and knowledge graphs to perform complex tasks such as data analysis, decision-making, medical assistance (e.g., dental diagnosis), and claims processing automation.


Key Responsibilities:
Design and implement agentic workflows using LangChain and LangGraph.

Build and deploy autonomous agents capable of task planning, execution, and collaboration.

Integrate external tools, APIs, and vector databases (e.g., Pinecone, FAISS, Chroma).

Optimize agent memory, tool use, and feedback loops for improved reasoning.

Create modular, scalable, and well-documented Python code.

I'm looking for a skilled professional to develop a hybrid model using GANs and CNNs for the localization of spliced natural images. The primary objective of this project is to enhance the accuracy of image localization.

Key Requirements:
- Utilize existing dataset for training
- Focus on natural images
- Employ a hybrid GAN-CNN approach

Ideal Skills and Experience:
- Expertise in GANs and CNNs
- Experience with image localization tasks
- Proficient in handling and training on custom datasets
- Strong background in computer vision and deep learning

Please provide examples of previous work and relevant qualifications.

Skills: Python, Image Processing, MATLAB, Convolutional Neural Network

I'm looking for a skilled professional to develop a hybrid model using GANs and CNNs for the localization of spliced natural images. The primary objective of this project is to enhance the accuracy of image localization.

Key Requirements:
- Utilize existing dataset for training
- Focus on natural images
- Employ a hybrid GAN-CNN approach

Ideal Skills and Experience:
- Expertise in GANs and CNNs
- Experience with image localization tasks
- Proficient in handling and training on custom datasets
- Strong background in computer vision and deep learning

Please provide examples of previous work and relevant qualifications.

Skills: Python, Image Processing, MATLAB, Convolutional Neural Network

Computer Vision Specialist Engineer – Garment & Human Segmentation (Deep Learning, PyTorch)

Project Overview
We’re building next-generation AI tools for the fashion industry and need a Computer Vision Specialist Engineer to develop precise garment and human segmentation models. Your work will power core features in our apps, delivering seamless digital try-ons, outfit analysis, and interactive experiences.

Key Responsibilities

Architect, train and deploy deep learning models for semantic & instance segmentation of garments and people

Create custom networks tuned for clothing attributes (e.g. fabric type, cut, drape) and human body parts under varied poses and lighting

Optimize models for low-latency, high-accuracy inference in production (mobile/web)

Lead data pipeline: collect images/videos, define annotation strategy, preprocess inputs, validate labels

Integrate vision modules with backend services and front-end clients in collaboration with product, design, and engineering teams

Produce clear technical documentation and present model results to stakeholders

Required Skills & Experience

Master’s in CS, EE, Applied Math, or related field

2+ years building and shipping computer vision solutions

Expert in segmentation architectures (Mask R-CNN, U-Net, DeepLab) with demonstrable projects on garments or fine-grained part recognition

Strong Python proficiency and hands-on experience with PyTorch or TensorFlow

Solid command of OpenCV and classic vision methods (edge detection, morphological ops)

Track record of tuning models for speed and resource constraints

Bonus Qualifications

Open-source contributions or publications in CV conferences (CVPR, ICCV, ECCV)

Familiarity with ComfyUI for UI-driven model workflows

Commodity Trading Coursework Report – CFD Simulation Analysis Project

Description:
I’m seeking a highly skilled individual with strong experience in commodity markets, derivatives trading (especially CFDs), and academic writing to assist with a project. The task involves completing a trading simulation using a demo account and writing a final report based on strategy, execution, risk management, and performance.

Project Details:

Objective: Simulate trades in commodity CFDs (energy, metals, agriculture) and produce a comprehensive final report (max 20 pages).
Platform: CMC Markets or IG UK (demo account).
Portfolio: Minimum 3 contracts across energy, metals, and agriculture (total 4–5 recommended).
Report Requirements:

Trading strategy (fundamental/technical/sentiment/scalping/etc.)
Market analysis with relevant data (news, forecasts, price trends)
Risk management approach (stop-loss, position sizing, allocation)
Trade execution commentary (rationale, performance)
Screenshots of demo account balance and trades (opening/closing)
Final reflections and learnings

What I’m Looking For:

Background in finance, trading, or commodity markets
Comfortable using demo trading platforms
Capable of writing clear, structured academic reports
Familiar with tools like Bloomberg, Eikon, or alternative data sources
Able to complete the project by early July 2025

Deliverable:
One well-structured PDF report (max 20 pages), with charts, screenshots, references, and sound analysis.

Timeline:
Start ASAP. Trading must be completed by 4 July 2025. Final report due by 6 July 2025.

Budget: Open to reasonable offers based on quality and experience.

To Apply:
Please share:

Your experience in commodity or CFD trading
Any similar academic/trading reports you've done
Which platforms/tools you’re comfortable using

I'm looking for a skilled AI Engineer to lead the end-to-end integration of artificial intelligence capabilities into our existing project. The ideal candidate will have hands-on experience connecting AI models (such as GPT, Hugging Face models) with web applications, APIs, or backend services.

I'm looking for a skilled developer to build a prototype of an Adversarial Role Chatbot System designed to support academic writing and critical thinking. The goal is to simulate critical dialogue, where different AI personas challenge, question, or critique user input, promoting deeper reasoning and reducing hallucinated or weak claims.



The system draws from the logic of existing tools like IBM’s Project Debater, DebateDevil, YesChat bots, and DebunkBot. It should allow users to engage with toggleable roles such as a Devil’s Advocate, Socratic Tutor, or Ideological Opponent. These agents should respond in-character, question user assumptions, cite counterarguments, and, optionally, reference external sources like PubMed or Semantic Scholar.



We have already detailed internal documentation outlining relevant prior work, strategies, and tool suggestions. Now we’re looking for the right developer to bring this vision to life.



✅ What the prototype must demonstrate:
- Role-based conversational agents (via prompt engineering or multi-agent setup)
- Toggleable adversarial personas with different tones (e.g., skeptic, critic, ideologue)
- Optional integration of citation support from tools like Elicit or PubMed APIs
- Option for user control over tone (e.g., slider for mild vs aggressive critique)
- UI/logic for persona switching and summarizing debates



💼 Ideal Experience:
- Python development with LLMs (OpenAI, Claude, DeepSeek, etc.)
- Experience with adversarial prompting, multi-agent systems, or dialogue orchestration
- Familiarity with LangChain, CrewAI, or custom agent frameworks
- Bonus: knowledge of academic reasoning, critical thinking models, or Socratic tutoring systems


🌍 Location:
Our team is based in North America (PDT) and Europe (CET). Please apply only if you're located in one of these or neighboring time zones to ensure good collaboration.


Hourly rate to be discussed, large range was applied.

mNDA required.

️ What You Need to Do:
If you believe you are a good fit for this role, apply and make sure to include the word "persona" in your application. This will let us know that you actually read the whole job posting

We’re looking for an experienced ML engineer to fine-tune the Qwen2.5-VL vision-language model using a custom dataset of ~5,000 annotated images. The task involves anomaly detection based on object placement.

⸻

✅ Requirements:
• Deep experience fine-tuning Qwen2.5-VL
• Proficiency with LoRA, PEFT, Hugging Face, and PyTorch
• Familiarity with multimodal instruction tuning
• Ability to run training on RunPod, Vertex AI, or similar A100 environment

⸻

🔧 Scope:
• Prepare training pipeline (JSONL format)
• Fine-tune and evaluate model
• Deliver model + simple inference script

Quick turnaround expected (project size is manageable).
If you’ve fine-tuned Qwen2.5-VL before — let’s talk.

• Describe your recent experience with similar projects

We're building the next generation of systematic trading infrastructure and looking for an experienced developer who's passionate about financial technology and thrives in a fast-paced startup environment.

The Opportunity:
This is a hands-on role where you'll architect and build our entire live trading execution platform from the ground up. You'll have significant technical ownership and the freedom to make key architectural decisions that will shape our trading infrastructure for years to come.

What You'll Build:

Design and implement our core trading execution engine
Build integrations with brokerage platforms and market data providers
Create automated execution systems with sophisticated risk controls
Develop real-time monitoring and alerting capabilities
Build position management and reconciliation systems

What We're Looking For:

Proven Track Record: Experience building and deploying live trading systems in production
Technical Expertise: Strong background in Python, C++, or Rust with real-time systems experience
Startup Mindset: Someone who enjoys wearing multiple hats and moving fast in an entrepreneurial environment
Ownership Mentality: You take pride in building robust, scalable systems that perform under pressure

Why This Role:

High Impact: Your work directly influences trading performance and company success
Technical Freedom: Make architectural decisions and choose the best tools for the job
Growth Potential: Strong performers will be considered for full-time positions as we scale
Learning Opportunity: Work alongside experienced traders and technologists in a collaborative environment

Path to Full-Time: Exceptional candidates who demonstrate strong technical skills and cultural fit will have the opportunity to transition into a permanent role with equity participation as we grow the team.

• Describe your recent experience with similar projects
• What techniques would you use to clean a data set?
• How do you deal with outliers or missing values in a dataset?

We are building an AI Interview Assistant that conducts intelligent, automated interviews using OpenAI's advanced language models. The system will simulate a conversational interviewer, asking relevant questions based on the job role, analyzing candidate responses in real time, and generating structured feedback for recruiters.

Key Features:

- Dynamic Interview Generation: Automatically generate tailored interview questions based on job titles or provided prompts.

- Conversational AI: Use OpenAI (GPT-4 or GPT-4o) to simulate natural, responsive dialogue during interviews.

- Real-Time Response Analysis: Evaluate candidate answers for clarity, depth, relevance, and soft skills.

- Interview Summaries: Generate concise reports highlighting candidate strengths, potential concerns, and overall impressions.

Description:
I'm building a machine learning system to detect and match tigers in camera images using their stripe patterns. The model should identify whether a tiger already exists in the dataset and return a confidence score, plus optionally show the closest matches.

🔧 Key Requirements:
Build a Siamese CNN model in Google Colab

Use tiger images for individual identification (not just detection)

Return confidence score for each match

Support auto-updating datasets without full retraining

Deploy as a cloud API (Google Cloud or AWS)

Modular, maintainable code

(Optional) Set up a basic web UI later

🧠 Ideal Skills:
CNNs, image classification, pattern recognition

Experience with Siamese/Triplet Networks

Python, TensorFlow/PyTorch, API deployment

Familiarity with animal datasets is a plus

I have two research papers, one which discusses Diffusion Model Sampling with Neural Operators and another which discusses Continuous Time Consistency Models. Need someone to read through these papers and completely reproduce the code/results. A strong mathematical background is required in Differential Equations, Linear Algebra, and Statistics. More info will be provided after contact. This project has many extention opportunities. This is a challenging project so I will only select the top candidates. I also need this done ASAP. Success in this job will lead to a long-time collaboration. Any proposals written by AI will be completely disregarded.

Paper 1: https://arxiv.org/pdf/2211.13449 (github: https://github.com/neuraloperator/DSNO)
Paper 2: https://arxiv.org/pdf/2410.11081 (unofficial git: https://github.com/xandergos/sCM-mnist)
More Background for Paper 2: https://arxiv.org/pdf/2303.01469

• What's your education experience and background in math and computer science?
• To the best of your understanding explain to me what a neural operator or consistency model is.
• Describe your recent experience with similar projects

AI SaaS for Dental Implant Diagnosis Using DICOM CT Scans

⸻

Project Overview:

We are building a web-based SaaS platform that uses Artificial Intelligence to analyze dental CT scans (DICOM format) and support implantology diagnosis.

The platform should:
• Allow users (dentists/clinics) to upload DICOM (.dcm) files from dental CT scans.
• Use AI to detect caries, cysts, and available bone volume based on visual patterns (color and location).
• Enable the dentist to simulate dental implant placement in a 3D environment.
• Be trained and validated with the support of a professional implantologist, who will provide labeled data.

This project is in its early stages and will start with an MVP.

⸻

Technical Requirements:

Frontend:
• React.js (preferred)
• Integration with OHIF Viewer or Cornerstone.js for DICOM rendering
• 3D implant simulation (Three.js or similar)

Backend:
• Python + FastAPI or Flask
• File upload system (DICOM)
• API to trigger inference from trained AI model

AI/ML:
• Medical image segmentation using PyTorch or TensorFlow
• Preferred model: nnU-Net or MONAI
• Input: DICOM or preprocessed volume data
• Output: Segmented zones (caries, cysts, bone availability)

Data Format:
• Input: .dcm (DICOM format)
• Annotations: Provided by an implantologist using 3D Slicer or similar tools

Deployment:
• Dockerized backend
• Cloud hosting (AWS/GCP)
• DICOM file storage (must be secure and GDPR-compliant)

⸻

Compliance & Ethics:
• All patient data must be anonymized.
• The platform must comply with GDPR (Europe) or HIPAA (USA) requirements for medical data privacy.

⸻
Deliverables:
1. Upload and preview CT scans in a web viewer.
2. AI model trained on provided labeled data (start with 5–10 scans).
3. API to return predictions/segmentations.
4. 3D interactive simulation of implant placement.
5. (Optional) Dashboard to manage patients and results.

⸻

Ideal Candidate:
• Experience with DICOM files and medical imaging
• Background in AI/ML (especially CNNs, segmentation models)
• Frontend skills for visualization and UX
• Familiar with GDPR/HIPAA compliance

⸻

To Apply:

Please include:
• Relevant experience with DICOM or dental/medical imaging.
• Links to similar projects or GitHub repos (if available).
• Estimated timeline for MVP delivery.
• Your suggested tech stack (if different from above).

Skills: Python, Frontend Development, Deep Learning, 3D Visualization

I am conducting a project focused on Natural Language Processing (NLP) for Arabic-language user reviews from banking mobile applications in the Middle East.

The dataset is already collected and cleaned. I now need a qualified NLP freelancer to deliver a complete solution including both Sentiment Analysis and Topic Detection using both traditional ML techniques and modern LLMs.

🔧 Responsibilities:

🟢 Sentiment Analysis:
- Implement traditional ML models: Naive Bayes, SVM, Logistic Regression, Random Forest

- Implement LLM/Deep Learning models: AraBERT, MARBERT ، GPT-3.5 / GPT-4 (if available via API)

🟢 Topic Detection:
- Apply unsupervised ML techniques: LDA (Latent Dirichlet Allocation)


📊 Evaluation:
- Provide performance metrics: Accuracy, Precision, Recall, F1 (for sentiment) Coherence score / qualitative interpretation (for topics)/Visualizations and comparison charts for all models

✅ Requirements:
- Proven experience in Arabic NLP tasks

- Strong knowledge of: scikit-learn ، HuggingFace Transformers ، Arabic LLMs (AraBERT, MARBERT, etc.)، Topic Modeling tools (LDA, BERTopic)

- Ability to write modular, reusable, and documented Python code

Fluent in Arabic is strongly preferred

📁 Deliverables:
End-to-end implementation of:

- Sentiment Analysis (ML + LLM)

- Topic Detection (LDA + BERTopic)

- All code with clear documentation

- Output files (CSV/Excel)

- Evaluation results with visual comparisons

We are looking for an experienced computer vision or machine learning engineer to develop a robust real-time facial recognition system. The system should be capable of detecting faces, verifying identities through face matching, and optionally implementing liveness detection to prevent spoofing (e.g., photos or video replays).

⸻

📌 Scope of Work:
• Build a pipeline that performs:
• Face detection (using Haar, DNN, or modern detectors like RetinaFace, BlazeFace)
• Face embedding extraction (e.g., with FaceNet, ArcFace, or Dlib)
• Face matching/verification
• (Optional but preferred) Implement liveness detection using eye-blink detection, head movement, or frame-based anti-spoof techniques
• Build a small demo (webcam/streamlit/web UI) to test system performance
• Create a simple enrollment/registration pipeline to save embeddings to a local DB or vector store
• Support at least 90%+ accuracy under reasonable lighting conditions

⸻

✅ Deliverables:
• Source code and model files (preferably open-source or pretrained)
• Working demo (desktop or web-based)
• Documentation and installation instructions
• Suggestions for deployment or optimization if needed (e.g., ONNX, quantization)

⸻

💡 Skills Required:
• Python (OpenCV, face_recognition, Mediapipe, or similar)
• Deep learning frameworks (TensorFlow or PyTorch)
• Facial embeddings and similarity metrics (cosine, Euclidean)
• Experience with real-time video processing
• (Bonus) Knowledge of liveness detection and anti-spoofing

We are seeking an experienced professional to assist with takeoffs using the stack method. The ideal candidate will have a strong understanding of takeoff procedures and be proficient in interpreting construction plans. You will work closely with our team to ensure accurate measurements and estimates are provided in a timely manner. If you have a proven track record of successful takeoff projects and can demonstrate your expertise, we would love to hear from you.

We have a client in the Testing, Inspection and Certification (TIC) sector, looking to implement an AI-enabled platform for the purpose of training and certifying technicians

AI and GenAI can modernize legacy training materials and enhance learner engagement through interactive content, improving retention and knowledge transfer. AI / Gen AI can generate test questions and real-world scenarios aligned with field applications, identify completed modules, and tailor learning to focus on remaining skill areas. Additionally, GenAI can adapt to individual learning pace and respond to questions using both course content and supplemental materials, such as OEM manuals and technical guides.

Business Problem: currently, it takes the company 6 months (minimum) to train a technician because the only way to do so today is to assign a billable/trained technician with a trainee. When they do that, they have the cost of labor and travel for both, and of course the opportunity cost of not engaging the technician in billable client work. The hypothesis is that if we can reduce the training cycle time in ½ or even take it down to 2 months, then the impact would be material to the business. The issue is getting all the paper, manuals and materials and other training items into a method that allows us to get the information to the technicians. Goal is to hire 10 technicians/year in Canada, and double that # in the US…and of course train the new techs. There are currently 150 technicians in Canada and 100 in the US.

Phase 1: Scoping of the Project
• Define Scope & Success Criteria: Establish clear objectives and measurable outcomes for the training program and evaluate data on hand and expected data requirements (i.e. training materials, training requirements specifications, training paths / certification requirements, technician training history, OEM manuals and materials, etc.)
• Process Design Enhancements: Identify opportunities for streamlining workflows and knowledge delivery
• Solution Mockup: Create preliminary interface designs and interaction models and a structure for Phase 2 which is the full implementation.
• Implementation Planning: Develop timeline, budget estimates, and buy vs. build recommendations

Phase 2: Full Implementation
• Following successful completion of Phase 1, provide project management oversight for the implementation of the field technician training solution across the company

Potential Approach:
• Start with a proof of concept for one line of business with this client
• The solution should integrate with the company’s operating database to have access to all the customer and other transactional data
• Should establish an LMS platform for technician training, testing and certification, to track status of technician training, with all associated metrics
• Produce training and certification documentation for basic lift inspection services
o Basic and specialty equipment
o Equipment that IIA has inspected before that is part of an annual inspection
o Pictures and design documentation for the actual equipment that details the inspection steps that would generate the highest degree of deficiencies and preventative maintenance services to perform, which translates into revenue dollars
• Power BI plugged into the LMS that can produce reports on the status and efficiency of the training and certification program
• Ultimately want this to be an operating platform for the company and other companies within the business portfolio

Job Title:
GoHighLevel + Synthflow AI Agent Builder for Local Business Automation

Job Description:


I'm looking for a skilled freelancer who can help build and deploy AI-powered voice or chat agents connected to GHL calendars, pipelines, and workflows. You'll also help configure GHL automations and ensure smooth handoff between AI, CRM, and follow-up sequences.

Responsibilities:

- Build and train Synthflow agents (inbound and/or outbound)
- Create workflows in GoHighLevel (form intake, tagging, SMS follow-up)
- Connect Synthflow to GHL using webhooks or direct integrations
- Set up GHL calendars, pipelines, and appointment logic
- QA all flows and ensure leads are routed correctly
- Build clean fallback logic and objection handling paths
- Optionally assist in designing agent avatars and branding

Required Skills:

- GoHighLevel CRM expertise (funnels, automations, calendars)
- Synthflow experience (or similar tools like Twilio Studio, Voiceflow, Chatbase, etc.)
- Webhooks, integrations (Zapier or native)
- Strong logic-building and flow structuring
- Fluent English (written and spoken)

Bonus Skills (Not Required, But a Plus):

- Experience building AI sales/chat agents
- Prompt engineering (ChatGPT, Claude)
- Graphic design for agent avatars or landing pages (Canva, Leonardo.ai)


How I Work:

I move fast, communicate clearly, and give you creative control — I just need a reliable builder who can execute cleanly and meet deadlines. I prefer Loom videos for feedback and use GHL and Notion to manage projects.

Project Scope:

This is a medium-sized project with the potential for long-term or ongoing work. You’ll start with one agent and booking system — and if all goes well, we’ll scale into multiple clients and systems.

Next Steps:

Please apply with:
- A brief summary of your experience with GHL and AI agents
- Links to past work or demo flows (if available)
- Your preferred timezone and availability

Let’s build something amazing.

What You'll Be Doing

Design, develop, and deploy AI/ML solutions using AWS, with a focus on Amazon Bedrock, Sagemaker, and related AI services, ensuring high performance and scalability.
Implement and optimize Retrieval-Augmented Generation (RAG) systems to unlock enhanced AI capabilities for various client applications.
Implement AI-powered document processing and workflow automation, demonstrating your knowledge in automating structured and unstructured data pipelines
Fine-tune large language models (LLMs) for specific business use cases, tailoring models for optimal performance and accuracy within the client's domain.
Build and optimize agentic workflows to enable dynamic, task-driven AI agents, streamlining complex processes and automating intelligent decision-making.
Architect scalable, secure, and cost-effective AI infrastructure on AWS, ensuring compliance with best practices and optimizing resource utilization.
Work closely with clients to understand their requirements and deliver tailored AI solutions that directly address their business challenges.
Stay up to date with the latest trends and advancements in AI/ML and AWS technologies, continuously learning and applying new knowledge to improve our solutions.


What We Need To See
Strong understanding of the ML fundamentals and experience with designing end-to-end ML pipelines, from data ingestion to model deployment.
Solid knowledge of Retrieval-Augmented Generation (RAG), document processing, LLM fine-tuning techniques, including practical experience in implementing and optimizing these technologies.
Experience with text/image embedding and working with vector databases to store vectors such as Pinecone, Weviate or Opensearch
Experience in AWS infrastructure and services, including Lambda, Step Functions, ECS, SageMaker, Bedrock, Opensearch, showcasing your ability to build and manage complex cloud environments.
Experience with libraries such as scikit-learn, statsmodels, PyTorch, Langchain, LangGraph, Langfuse, demonstrating your proficiency in utilizing these tools to develop and deploy AI/ML solutions.
Experience in Python
Experience working in a client-facing role, managing technical discussions, and ensuring client satisfaction, demonstrating your ability to build strong client relationships and deliver high-quality solutions.

• Describe your recent experience with similar projects

My name is Saravanan Vellaiyan, and I am the owner of a civil engineering firm based in Ontario, Canada. We have a strong reputation for accuracy and efficiency, and now we're looking to build a cutting-edge tool to further enhance our services.

We are seeking an expert-level AI/Machine Learning engineer to partner with us in developing a proof-of-concept and eventual MVP (Minimum Viable Product) for an AI-powered tool that can automate key parts of our services process.

The core challenge is to build a system that can analyze Construction Drawings (in PDF or DWG format) to identify, classify, and quantify specific components and their attributes from these documents.
This is a unique opportunity to apply your AI skills to a real-world problem and build a groundbreaking product in the construction technology space.

I am a subject matter expert in Civil Engineering and will be working very closely with you, providing the necessary data, industry knowledge, and feedback to ensure this project is a success. We are looking for a strategic partner, not just a coder.

We will start with a paid, smaller-scope discovery phase to define the project and validate the approach before committing to the full MVP development.

We are excited to find the right partner to help bring this vision to life!


Scope of Work & Deliverables:
We will approach this in two phases:

Phase 1: Paid Discovery & Feasibility Study (Initial Milestone)
Analyze a sample set of our construction drawings to assess complexity.
Define the most effective technical approach (e.g., which models to use, data labeling strategy).
Create a detailed technical specification and roadmap for the MVP.
Provide a reliable time and cost estimate for the MVP development.

Phase 2: MVP Development
Develop and train a computer vision model to accurately detect and classify the components on structural drawings.
Build a simple, functional web interface where a user can upload a drawing file.
The application should process the drawing and output a structured data file (e.g., CSV or Excel, PDFs)
The initial MVP will focus on a limited set of common rebar types as a proof-of-concept.
Required Skills & Experience:
Please only apply if you have proven, hands-on experience in the following areas:

Technical Skills:
Expert-level Python
Deep Learning & Computer Vision: Extensive experience with object detection/segmentation models (e.g., YOLO, Mask R-CNN) is essential.
ML Frameworks: Mastery of TensorFlow or PyTorch.
Image Processing Libraries: Strong skills with OpenCV.
Data Handling: Proficiency with Pandas and NumPy.
Experience with reading/parsing PDF or CAD files is a major plus.
Experience deploying ML models to a cloud service (AWS, Google Cloud, or Azure).
Professional Skills:

Excellent communication skills and fluency in English.
Ability to explain complex technical concepts to a non-technical person.
A problem-solving mindset and a focus on delivering business value.
Verifiable portfolio of past AI/ML projects that have been successfully deployed.

• 1. Please describe a past computer vision project you worked on where you had to extract information from images or documents. What was the challenge, what was your approach, and what was the outcome?
• 2. Based on the project description, what do you foresee as the biggest technical challenge in building this AI tool?
• 3. Please share a link to your portfolio or specific case studies of relevant work (AI, ML, or Computer Vision projects).
• 4. Are you available during business hours in the Eastern Time Zone (EDT) for brief weekly check-in calls?

We are seeking a professional collaborator who can take full responsibility for enhancing the codebase and writing a comprehensive research paper that meets international academic publication standards. Develop a hierarchical deep learning framework that integrates multiple AI models to analyze heterogeneous medical data, with a particular focus on blood test results and genomic data related to blood cancers. The models involved include Transformers, CNNs, and reinforcement learning or grid search, and the architecture should support both local and global feature extraction using attention mechanisms and residual connections. This project aims to deliver not only high classification accuracy but also meaningful interpretability to assist medical professionals in making accurate and timely diagnostic decisions. Key Objectives 1. Develop a robust and scalable deep learning framework that processes multi-modal medical data. 2. Enhance feature extraction using attention layers and residual connections for better handling of complex biomedical signals. 3. Improve classification performance via a confusion strategy combining predictions from multiple models. 4. Evaluate the performance of the proposed system against traditional ML algorithms and existing diagnostic approaches. 5. Integrate multiple Explainable AI (XAI) techniques to ensure model interpretability, including visual explanations for medical professionals. 6. Apply the framework to two datasets that we will provide. 7. Perform 5-fold cross-validation and standard evaluation (training/validation/testing) on both datasets, and compare results in the paper. 8. Apply Grid Search for model optimization and maintain the use of confusion matrices throughout for performance evaluation. 9. Achieve accuracy improvements, particularly on the both dataset, by any suitable means—including new models or architectural changes. ⸻ Explainable AI (XAI) Component • Implement multiple XAI tools ( SHAP, LIME, Grad-CAM, Integrated Gradients). • Visualize the decision-making process of the deep learning models across datasets. • Include a comparative evaluation of the effectiveness of each XAI method. Research Paper Requirements • The final paper must be comprehensive and academic-grade, intended for submission to a high-ranking peer-reviewed journal. • Length: more than 35 pages to reflect the depth and complexity of the work. • Include all standard academic sections: Abstract, Introduction, Related Work, Methods, Experiments, Results, Discussion, Conclusion, and References. • A literature review of at least 25 academic studies is required. • The paper must undergo a plagiarism check using iThenticate and turinitin. We will provide the report, and the final version must have a similarity index below 9%. • Integrate comparative analysis across datasets, models, and XAI tools. Note: we have code but need enhance and improve on it. I need publish this paper in ISI Q1.

Skills: Research, Technical Writing, Article Rewriting, Artificial Intelligence, Deep Learning

Full AI Sky Replacement System for Server-Side Implementation (Photoshop Quality)"

--- Overview:

I am looking for an experienced computer vision engineer or AI developer to implement a fully functional, high-quality sky replacement system that runs on a Linux server (ideally FastAPI-compatible). The system should deliver results comparable to Photoshop or Luminar AI, with proper sky detection, segmentation, blending, and lighting correction.


--- Deliverables:

1. Sky segmentation model (pretrained or custom trained)


2. Foreground-background separation with edge handling (trees, hair, etc.)


3. Sky blending & tone matching algorithm


4. Batch API endpoint (e.g., FastAPI) to upload photo + sky and get result


5. Post-processing pipeline (blurring, feathering, color correction)


6. Optional: Skin/human area preservation


7. Complete source code, environment setup, and deployment guide

--- Preferred Technologies:

Python (OpenCV, PyTorch/TensorFlow, NumPy, scikit-image)

FastAPI (for API)

Any efficient model for segmentation: U^2-Net, DeepLabV3+, MODNet, SkyAR, etc.

--- Sample Input/Output:

Please include examples where:

The sky overlaps with trees or hair

There is backlighting or exposure differences

Overcast or gray sky needs vibrant replacement

--- Constraints:

Must run efficiently on GPU-enabled server or optimized CPU fallback

Output should be high-resolution
Open-source base preferred (or license-cleared)

No third-party API (must run locally)

--- Timeline:

Delivery within 1 week including revisions.

Skills: Linux, Artificial Intelligence, Image Processing, Deep Learning, FastAPI