We’re hiring a Senior AI Integration Technical Lead (hands-on) to build and deploy an AI-enabled operations layer that materially improves agent efficiency, knowledge accuracy, and Tier-1 customer support resolution for a regulated financial services workflow. This is not a research project. We need a builder who can own the technical execution end-to-end with minimal coaching. Our team will lead project management and client coordination; you will deliver the majority of the technical work. What You’ll Build: You will deliver three production-ready tools built around an existing operations platform (no deep core platform changes required in Phase 1): 1) Internal Knowledge Copilot (Program/Policy + Ops Intelligence): Build a structured, searchable, version-controlled knowledge base from internal program/policy documents. A secure internal web app for instant “rule lookup” and Q&A 2) Outbound Message Review + Validation Tool: This ingests daily inbound/outbound communications (CSV/export/API). Applies rules and context checks (document status, recent updates, review flags, etc.). Produces an auditable output: send / suppress / delay / revise 3) Human-First Chatbot(s): A.) Customer-facing chatbot for FAQ and basic workflow guidance. Must have clear escalation option to live human support (no “call deflection walls”) B.) Internal agent-facing chatbot that leverages the same knowledge base as the copilot --------------------------------- Required Deliverables: - Architecture diagram + short design doc - Auth approach (SSO preferred) + secrets management plan - Logging/audit design: prompt versions, source citations, user actions, outputs, and decision records Knowledge System: - Ingestion pipeline (upload/import → chunking → metadata → versioning) - Retrieval layer (vector search / hybrid search) configured for grounded answers Apps & Services - Internal copilot web app (search/Q&A, citations, training mode) - Message review tool (ingest → checks → outputs → audit trail) - Chatbot backend service + embeddable widget (customer + internal) -------------------------------------- Preferred Stack (flexible if you justify better) 1. Azure/OpenAI-compatible LLM stack (Azure OpenAI preferred) 2. Vector retrieval (Azure AI Search preferred, or equivalent) 3. Backend: Python (FastAPI) or Node.js 4. Frontend: React/Next.js (or equivalent) 5. Auth/SSO: Entra ID / OAuth2 / SAML as applicable 6. Docker + CI/CD preferred -------------------------------------------- Must-Have Experience: - You have shipped production RAG / knowledge assistants (not demos) - You have built secure enterprise systems (regulated environments strongly preferred) You understand and can implement: 1.) audit logs / traceability / prompt and doc versioning 2.) safe prompt patterns and guardrails (no leaking sensitive info) - You can independently drive architecture → build → deploy → documentation ------------------------------- Nice-to-Have: - Azure infrastructure depth (Key Vault, App Service/Container Apps, private networking) - Experience integrating with email/SMS platforms, ticketing systems, or CRM tools ------------------------------------- Engagement: Hourly or fixed monthly retainer 1.) 20–40 hrs/week, flexibility as the project demands change. 2.) You will work directly with our management team and provide weekly progress updates --------------------------------- How to Apply (Required) Please include: 2–3 examples of similar systems you built (links, brief writeups, or anonymized diagrams) The technical stack you recommend here and why. Feel free to ask as many questions as needed.