Full-Stack AI Developer for Document-Based AI Chatbot

We’re looking for an experienced Full-Stack AI Developer to build an AI chatbot that can answer questions accurately based on our internal documents and uploaded PDF files. The application will be integrated into an existing modern web application and should use a retrieval-based AI architecture so that responses are grounded in the provided documents rather than relying only on the language model’s general knowledge. Users should be able to upload documents, ask questions in a conversational interface, and receive relevant answers along with clear references to the source material used. You will be responsible for both the frontend chatbot experience and the backend AI pipeline. This includes processing uploaded documents, extracting and cleaning text, breaking content into useful chunks, generating embeddings, storing them in a vector-based search system, retrieving the most relevant context for each question, and passing that context to the language model to generate the final response. The frontend should be built with a modern React-based framework using TypeScript, while the backend can use either JavaScript/TypeScript or Python depending on the architecture you recommend. Experience building REST APIs, document-processing pipelines, vector search, retrieval-augmented generation, and production AI integrations is important. We also want the chatbot to display source references so users can understand where an answer came from. The final application should be cleanly structured, maintainable, documented, and ready for deployment to a modern cloud hosting environment. We are particularly interested in developers who have previously worked on AI chatbots, document question-answering systems, internal knowledge assistants, or other AI SaaS applications. In your proposal, please briefly explain a similar project you have worked on, how you approached document retrieval and grounding, and what technical decisions you made to keep the responses accurate and reliable. We will begin with a small paid test task before moving into the full project. The test will allow us to evaluate code quality, communication, and your approach to AI retrieval workflows. If the initial work goes well, there will be opportunities for additional AI features and longer-term development work.