Senior AI Engineer for Enterprise RAG Chatbot (AWS, Vector DBs, Complex Data)

We are seeking a highly experienced AI engineer or small expert team to design and implement a custom, enterprise-grade conversational AI system for a client with a very complex software platform. The objective is to build a deeply trained, hyper-specific chatbot capable of reasoning across multiple internal knowledge sources and holding precise, context-aware conversations grounded strictly in the client’s proprietary data. This is not a generic chatbot or simple document Q&A project. We are looking for advanced architecture, thoughtful data ingestion strategies, and production-ready AI engineering. Core Objectives Build a chatbot that can answer highly specific, technical questions about the client’s software and internal operations Enable the system to reason across multiple heterogeneous data sources simultaneously Maintain accuracy, traceability, and contextual depth in responses Design for scalability, maintainability, and future data expansion Data Sources (Current & Planned) The system must ingest, normalize, and reason across data including but not limited to: Zendesk tickets (historical and ongoing) Confluence documentation and internal knowledge bases Emails (structured and unstructured) Slack messages and threads Product manuals and technical documentation Additional structured and unstructured datasets as needed Technical Expectations We are open to architectural recommendations, but experience with the following is highly relevant: Advanced RAG (Retrieval-Augmented Generation) architectures Vector databases (Pinecone, Weaviate, OpenSearch, FAISS, etc.) Embedding strategies, chunking, semantic search, and relevance tuning LLM orchestration and prompt engineering for multi-source reasoning Data ingestion pipelines for large-scale, continuously updated datasets Evaluation frameworks for accuracy, hallucination reduction, and grounding Security and access control considerations for enterprise data Some of the system is already built using AWS services (including Bedrock and S3), but we are not locked into a specific approach. Strong candidates should be comfortable: Extending the existing AWS-based architecture or Proposing a superior alternative if it materially improves results Ideal Candidate Profile Senior-level AI/ML engineer or AI solutions architect Demonstrated experience building production RAG systems Experience with enterprise knowledge systems (support tickets, internal docs, chat logs, etc.) Strong understanding of LLM limitations and mitigation strategies Comfortable making architectural decisions and defending them Clear communicator who can explain trade-offs and design rationale Deliverables System architecture design (diagram + explanation) Data ingestion and indexing pipelines Core chatbot implementation Deployment guidance and documentation Recommendations for ongoing optimization and scaling How to Apply Please include: A brief summary of relevant past projects (links or descriptions preferred) Your recommended technical approach at a high level Any experience with AWS Bedrock or similar LLM platforms Confirmation of availability and estimated timeline