We are a technology services firm supporting client projects in applied machine learning. We are looking for a senior machine learning engineer or applied scientist on contract to support model evaluation, training, and fine-tuning work. This is a hands-on role. We need someone with real machine learning fundamentals who has built, trained, and improved production models, rather than someone whose experience is primarily API integration. WHAT THE WORK INVOLVES - Designing and running model evaluation frameworks and benchmarks - Training and fine-tuning models for production use cases - Computer vision work including classification, detection, OCR, segmentation, and document or image understanding - Multimodal systems combining image and text or multiple data sources - Error analysis and iteration on underperforming models - Advising on dataset strategy, labeling approach, model selection, and experiment design BACKGROUND WE ARE LOOKING FOR - Machine Learning Engineer, Applied Scientist, or Research Engineer experience - Deep Python and strong PyTorch - Production experience with training pipelines and model evaluation - Ability to explain tradeoffs and recommendations clearly to client stakeholders ENGAGEMENT - Contract through Upwork with potential for ongoing work - Flexible hours with responsiveness expected during US business hours - Work performed under a subcontractor agreement covering confidentiality, IP assignment, and non-solicitation - We are only responding to independent contractors, not agencies. TO APPLY Send a short summary of your relevant work with links to GitHub, publications, or project writeups.

We are seeking a US-based computer vision and full stack developer to build a platform for sports card recognition. The project includes developing subscription management, dashboards, and user account features. The ideal candidate will have experience creating scalable applications and integrating computer vision capabilities into a user-friendly platform. Hiring: Computer Vision + Full Stack Developer for Sports Card Live Auction Overlay App (SaaS) 📌 Overview I’ve built an MVP of a real-time sports trading card scanning and comping overlay tool using Loveable.dev. The product helps buyers gain an edge during live auctions by instantly identifying cards and showing real-time market comps. Now I’m looking for a U.S.-based developer (or strong US-aligned freelancer) to take this from MVP → production SaaS. This is a subscription-based product, so I need someone who can help build something fast, accurate, scalable, and hard to replicate. 🧠 What the product does Users can: Capture or upload sports trading card images during live auctions (mobile + desktop) Instantly identify: Player Year / set Parallel / serial number Pull live market comps Display a real-time “buy / avoid / fair price” overlay The goal is speed + accuracy in live buying situations (seconds matter). ⚙️ What I already have MVP built in Loveable.dev Basic overlay + UI flow Initial comp logic concept Subscription idea (not yet fully implemented) 🛠️ What I need help building (Phase 1 → Scale) I’m looking for someone to help rebuild and harden the system into a real SaaS product: 1. Computer Vision / OCR Layer Card detection from images (mobile + desktop) OCR extraction (player name, set, serial numbers) Image recognition / matching to known cards Confidence scoring (very important — must avoid wrong matches) 2. Comp Engine (Core Value) Integrate or build system for: eBay sold listings 130point or similar comp sources Card Ladder / ALT-style pricing logic Return: last sale average comp trend direction liquidity estimate 3. Real-Time Overlay System Lightweight overlay that works during live auctions Low latency (fast lookup is critical) Works on mobile + desktop workflows 4. SaaS Infrastructure User accounts + authentication Subscription billing (Stripe) Usage tracking / rate limiting Admin dashboard 5. Scaling / Production Hardening API architecture improvements Database structure Performance optimization for real-time use Error handling for imperfect images 💡 Ideal candidate You should have experience with: Computer vision (OpenCV, YOLO, or similar) OCR pipelines AI image classification or similarity matching Full-stack SaaS development Stripe subscriptions API design (Node.js / Python / Next.js preferred) Huge plus if you have: Sports card / collectibles knowledge Experience with marketplaces or scraping pricing data Real-time / low-latency systems 🎯 Why this is interesting This is not a generic app. It’s: A real-time decision engine for high-value collectibles Built for a passionate, high-spend niche (sports cards) Subscription-based with strong monetization potential Designed for speed advantage in live auctions 📍 Requirements Must be U.S.-based (preferred for communication/time zone alignment) Must be able to work independently Must have strong GitHub/code examples Bonus if you’ve built AI or vision-based SaaS tools before 💰 Budget Open to: Hourly or fixed project 📩 To apply, please include: Relevant CV / GitHub Past AI / computer vision projects Any SaaS or startup experience Your approach to building a real-time image → comp system Availability per week

I am looking for an AI and machine learning developer to help create a basic image-recognition application for Decathlon. The goal of the project is to build a model that can identify the difference between running shoes and soccer cleats based on uploaded images. The developer should have experience with artificial intelligence, machine learning, computer vision, image processing, Python, and TensorFlow. The project would include helping develop and train the image model, testing its accuracy, and creating a basic prototype that could potentially be used in a mobile app or on a website. The main deliverables would be a trained image-recognition model, testing results, and a basic working prototype. I am looking for someone with relevant experience, strong communication skills, and the ability to complete the project within the expected timeline and budget.

We’re looking for an experienced artificial intelligence engineer to join the revolution, using deep learning, neuro-linguistic programming (NLP), computer vision, chatbots, and robotics to help us improve various business outcomes and drive innovation. The engineer will join a multidisciplinary team helping to shape our AI strategy and showcasing the potential for AI through early-stage solutions. This is an excellent opportunity to take advantage of emerging trends and technologies and make a real-world difference.

Seeking an experienced PyTorch engineer with strong knowledge of transformer internals to help implement a model-specific runtime intervention module within an existing supervisory architecture. The core architecture, evaluator, control logic, interfaces, and project structure are already implemented. This engagement focuses on implementing and validating the actuator layer that translates neutral control directives into model-specific residual-stream operations.

3rd Eye Robotics is looking for a senior, hands-on engineer to provide technical support and targeted development for **OmniVai®, our production edge AI/computer-vision platform for heavy equipment and industrial environments**. This is a flexible part-time contract role. **Typical workload is expected to be 5–10 hours/week, with some weeks requiring very little support and occasional periods up to 20 hours for development or platform migration work. Denver/Boulder-area candidates are strongly preferred.** The immediate need is someone who can become familiar with the existing production codebase and step in when technical issues arise, make configuration or deployment changes, and support continued commercial deployments. **Core experience:** * Strong **Golang and C/C++** * Production computer vision / object detection, including **YOLO** * NVIDIA **Jetson / CUDA / edge AI deployment** * Linux and embedded/edge systems * Real-time video and camera pipelines * Object tracking and depth/spatial perception * Ability to quickly understand and work independently in an existing production system **Near-term work includes:** * Porting our current edge software to the latest **Syslogic / NVIDIA Jetson Orin platform** * Supporting production deployments and troubleshooting when needed * Making configuration, software and model deployment changes * Maintaining the existing edge perception stack For the right engineer, there is also an opportunity to work with our platform team on **new perception-driven applications and rule logic** — for example, evaluating relationships between detected objects and site conditions, generating structured events, and sending those events to our cloud platform for operational insights. **Helpful experience:** * Luxonis OAK / DepthAI / stereo vision * Yocto / embedded Linux * Python and ML training/deployment pipelines * RTSP and IP cameras * OTA software/model deployment * Robotics, autonomous systems or heavy equipment * Distributed systems / cloud-connected edge devices * Android/Kotlin is a bonus, but not required OmniVai is already commercially deployed and includes multi-camera detection, temporal tracking, depth estimation, configurable application logic, field-data collection, cloud fleet management and continuous model improvement. We are looking for a **senior engineer who can provide reliable technical ownership without requiring a full-time role today**, with the potential for the relationship to expand over time.

We are looking for a skilled, hands-on AI Engineer to help us build and optimize our AI product. You will be responsible for designing the AI architecture, integrating modern LLMs/frameworks, and ensuring our AI pipeline runs efficiently, reliably, and accurately in production. Responsibilities Design, build, and deploy custom AI solutions (LLM integration, RAG, AI agents, or fine-tuning). Build robust prompt engineering pipelines, function-calling workflows, or structured output mechanisms. Implement vector databases (e.g., Pinecone, Weaviate, Qdrant, ChromaDB) for semantic search and retrieval. Optimize latency, API costs, and context window efficiency across LLM providers (OpenAI, Anthropic, open-source models). Connect AI models to backend services via REST APIs / webhooks. Implement evaluation metrics (hallucination detection, retrieval accuracy, output validation). Required Skills & Qualifications Languages: Python (strong expertise required), TypeScript/Node.js (a plus). AI / ML Tooling: LangChain, LlamaIndex, AutoGen, CrewAI, or direct SDK integrations (OpenAI, Anthropic, Hugging Face). Databases: Vector databases (Pinecone, Chroma, Qdrant, pgvector) + relational/NoSQL DBs. Deployment & Cloud: Docker, AWS / GCP / Azure, FastAPI / Flask, Serverless architectures. Core Concepts: In-depth understanding of Embeddings, RAG, Fine-Tuning, Function Calling, and Agentic Workflows. Preferred (Nice to Have) Experience deploying open-source models locally or on dedicated hardware (vLLM, Ollama, Hugging Face TGI). Experience with fine-tuning techniques (LoRA, QLoRA). Background in frontend AI UI integration (Vercel AI SDK, Streamlit, Gradio).

We are looking for an experienced Full-Stack AI Developer to build and improve modern AI-powered web applications. The ideal candidate should have strong experience in frontend and backend development, AI integrations, LLMs, LangChain, RAG pipelines, vector databases, APIs and scalable system architecture. Required Skills: * React.js, Next.js and TypeScript * Node.js and Python * LangChain, LLMs and RAG * OpenAI, Claude or similar AI APIs * Vector databases such as Pinecone, Qdrant or Weaviate * REST APIs, databases and third-party integrations * Strong knowledge of software architecture and clean code * Experience with Claude Code, Cursor or similar AI development tools * Excellent English communication and problem-solving skills Please share your GitHub profile, relevant AI projects and a short Loom video explaining one live Full-Stack AI application you have built.

## Job Title **Document AI / OCR App for Digitizing Magazine Images** ## Job Description I am looking for a developer to build a simple document-processing application for a large collection of magazine pages saved as PNG images. The application should process the images, extract the magazine text, identify important article information, and create organized digital files. The goal is to preserve the content while allowing the original image collection to be removed from storage. Magazine pages may include multiple columns, headings, captions, page numbers, advertisements, and images, so the solution should handle common magazine layouts. ## Required Features The application should: * Accept individual PNG images or folders containing many images. * Group multiple pages belonging to the same article when possible. * Extract the article text using OCR or a document AI service. * Preserve the correct reading order for multi-column layouts. * Identify: * Article title * Author * Publication date * Magazine or publication name, when available * Generate: * Full extracted text * A concise summary * A list of key points * Clearly mark information that could not be identified instead of inventing it. * Process large batches without requiring each image to be handled manually. * Save the results in an organized folder structure. Preferred output formats include Markdown, TXT, or JSON. A suggested article file might look like: ```markdown # Article Title **Author:** Author Name **Publication:** Magazine Name **Publication Date:** Month Day, Year **Source Images:** page_001.png, page_002.png ## Summary Brief summary of the article. ## Key Points - First important point - Second important point - Third important point ## Extracted Text Full article text in the correct reading order. ``` ## Deliverables * A working application or command-line tool; perferrablly a desktop application * Source code * Installation and setup instructions * Instructions for processing new folders of PNG images * Configurable output location and file format * Error logging for images that cannot be processed * A test run using a sample set of magazine images * Documentation of any external APIs, AI models, or ongoing usage costs ## Preferred Qualifications * Experience with OCR, document AI, or intelligent document processing * Experience handling multi-column magazine, newspaper, or scanned-document layouts * Python development experience * Experience with tools such as Tesseract, PaddleOCR, AWS Textract, Google Document AI, Azure Document Intelligence, or multimodal AI models * Experience producing structured metadata and AI-generated summaries * Ability to create a reliable batch-processing workflow ## When Applying Please include: 1. A brief description of your proposed technical approach. 2. Which OCR and AI tools you would use. 3. Whether the application would run locally, use cloud services, or use a combination of both. 4. Any estimated API or processing costs for a large image collection. 5. Examples of similar OCR or document-processing projects. 6. How you would handle articles that span multiple images or pages. 7. How you would verify the reading order and reduce OCR errors. Sample images may be sent upon request.

Hello! I’m looking for a thoughtful and reliable developer to create a simple Document AI application for a personal collection of articles, academic papers, and book chapters. The app should accept images containing text and use both **OCR and an LLM** to: 1. Extract the text from each image while preserving headings, paragraphs, and reading order as accurately as possible. 2. Organize the extracted text into logical sections. 3. Generate a concise summary for each section. 4. Generate an overall summary for the full article, paper, or chapter. 5. Save the original extracted text and summaries in an easy-to-use format, such as Markdown, TXT, or DOCX. The extracted text may later be used with text-to-speech software, but text-to-speech is **not part of this project**. The images may vary in quality and layout, so I would appreciate someone who can recommend an appropriate OCR approach and thoughtfully handle issues such as page order, columns, headings, footnotes, and repeated headers or page numbers. ## Preferred Qualifications * Experience with Python and OCR tools or services * Experience integrating LLM APIs * Familiarity with document structure, reading order, and long-document summarization * Ability to create a simple, approachable interface * Clear and patient communication * Respect for the privacy of uploaded documents ## Initial Deliverables * A working application that can process one document containing multiple images * Extracted and organized text * Section-level summaries * One complete document or chapter summary * Exportable output files * Basic setup and usage instructions * Source code I would be happy to provide several sample images so we can first confirm the OCR quality and discuss the best approach before building the full application. When applying, please share: * A brief explanation of how you would approach the project * Examples of similar OCR, document-processing, or LLM applications you have created * Which OCR and LLM technologies you would recommend * Whether you would suggest a desktop, local web, or cloud-based application * Any questions you have about the documents or desired output Thank you very much for taking the time to read this. I’m hoping to find someone who is careful, kind, and genuinely interested in making the application accurate and pleasant to use.