AI Knowledge Management Chatbot for Manufacturing Plant Documentation (RAG, Azure, Voice AI)

We are looking for an experienced AI and Azure solutions team or freelancer to design and build an AI powered Knowledge Management System for a manufacturing company operating multiple plants across India. Our current documentation, Standard Operating Procedures, maintenance protocols, and safety standards, is scattered across files and requires manual searching, which is slow and error prone. We want to replace this with an intelligent, RAG based chatbot that lets employees interact with plant documentation through natural language, both text and voice, in Hindi and English. Initial deployment will cover one plant location, with the architecture designed to scale later across our full footprint of multiple locations and plants, using access control and document segregation rather than separate deployments per plant. Scope of work: Ingest around 175 documents (PDF, Word, Excel) across Safety, SOP, Operations, and Maintenance categories. Use a document intelligence service to extract text, tables, images, diagrams, and charts, then chunk, embed, and index them for semantic search. Build a conversational chatbot on Azure OpenAI that answers queries with source references and can surface diagrams, charts, and tables directly from source documents, with links back to the original file. Integrate speech to text and text to speech so users can ask questions by voice in Hindi or English, with natural sounding responses. Build a web based user portal with SSO login, conversation history, and a document library, plus an admin portal with cost dashboards, usage monitoring, access management, and a kill switch for emergencies. Implement plant wise, role based access control so users only see documents relevant to their plant, with safety documents shared across all plants. Set up security and compliance: gateway level protection, rate limiting, OAuth2, content moderation, and PII and PHI masking. Set up monitoring and observability across the whole system. Optionally, generate short AI instructional videos, around 150 videos of 2 to 4 minutes each, from safety and maintenance SOPs to help field workers. Requirements: Strong experience with Azure AI Foundry, Azure OpenAI, Azure AI Search, and Azure Document Intelligence Experience building production RAG chatbot systems, not demos Experience with Azure AI Speech for voice interfaces Familiarity with role based access control and Entra ID SSO Bonus: experience with AI video generation tools such as Sora, Kling, or similar Project length: Multi month engagement, phased delivery Experience level: Expert Please share relevant past work on RAG chatbots or enterprise AI knowledge systems when applying.