AI Developer Needed to Build an AI-Powered University Document Analysis Assistant

AI Developer Needed to Build an AI-Powered University Document Analysis Assistant Project Overview We are looking for an experienced AI/LLM developer to build the first MVP of an AI Rector Assistant for a university. The goal is to develop an AI-powered assistant that analyzes Microsoft Word (DOCX) documents, compares them against a knowledge base of laws and university regulations, and generates analytical reports with compliance checks, risks, and recommendations. The assistant will initially work through Telegram (preferred over a web interface) to simplify deployment and user access. Scope of Work 1. Telegram Interface Develop a Telegram bot that allows users to: upload DOCX documents; receive document processing status; receive a structured AI-generated report; ask follow-up questions about the uploaded document; access previous analyses. 2. Document Processing The system should automatically extract: document type; author (if available); creation date; document purpose; document structure; key sections; key statements; references to laws and regulations. 3. AI Analysis The assistant should compare the uploaded document with a knowledge base containing: Laws of the Republic of Kazakhstan; Law on Education; Labour Code; Law on Science and Technological Policy; Ministry of Science and Higher Education orders; State Educational Standards; Professional Standards; University regulations; Rector's orders; Academic Council decisions; Internal policies and procedures. 4. Compliance Checking The AI should identify: applicable legal provisions; missing mandatory clauses; conflicts with legislation; outdated legal references; references to repealed regulations; duplicated content; legal and administrative risks. 5. AI Report Generation The final report should include: Executive Summary; Document Overview; Applicable Regulations; Compliance Assessment; Missing Requirements; Identified Risks; Recommendations; Overall Compliance Score. 6. AI Chat After analysis, users should be able to ask questions such as: Which regulations govern this issue? Does this document comply with current legislation? Which university policies should be updated? Are there similar internal documents? Which Ministry orders apply? The AI should provide answers with references to the relevant regulations whenever possible. 7. Analysis History The system should store: uploaded documents; generated reports; document versions; analysis history; user comments. Preferred Technical Stack We are open to suggestions, but we expect experience with technologies such as: Python LangChain or LlamaIndex OpenAI GPT-4.1 / GPT-5 (or equivalent LLM) RAG (Retrieval-Augmented Generation) Vector Database (Qdrant, Pinecone, Weaviate, Chroma, etc.) PostgreSQL Telegram Bot API Docker OCR (future phase) Expected AI Architecture The solution should follow a Retrieval-Augmented Generation (RAG) architecture. The AI should not rely only on the language model. Instead, it should: retrieve relevant legal documents; perform semantic search; compare document content with regulations; generate grounded responses; cite the relevant sources whenever possible. Deliverables Fully functional Telegram Bot AI document analysis pipeline RAG knowledge base Compliance checking module AI-generated analytical reports Source code Deployment documentation API documentation (if applicable) Nice to Have Experience with: Legal AI Document Intelligence University Management Systems Knowledge Graphs AI Agents Enterprise Search Microsoft Word document parsing Future Phases This project is the first stage of a larger AI platform for university leadership. Future modules may include: strategic planning; education trend analysis; AI adoption monitoring; university KPI analytics; economic forecasting; research analytics; executive decision support. We are looking for a long-term developer or development team to continue building the platform after the MVP.