Lead AI Architect & Delivery Manager

Role Summary

The Lead AI Architect & Delivery Manager will serve as the technical authority and engagement lead for the client's TPRM AI Assistant project, backfilling the departing AI Lead. This role is responsible for driving the technical strategy of the Snowflake-native RAG solution, orchestrating the offshore engineering team, interfacing directly with FCB leadership (ED&A, TPRO, InfoSec SMEs), and ensuring the solution meets its tight production launch milestone.

Key Responsibilities

• Technical & Strategic Leadership: Own the end-to-end architecture of the Snowflake Cortex RAG pipeline, ensuring alignment with Enterprise TPRM Standards 
• Client & Stakeholder Management: Act as the primary technical liaison to the client's business owners, AI leads, and InfoSec SMEs.
• UAT Execution & Triage: Lead User Acceptance Testing (UAT) sessions with InfoSec SMEs, converting user feedback into prioritized technical backlog items for the offshore pod.
• AI Governance & Compliance: Own the preparation of the Model Development Documentation (MDD) packet and drive AIEC-2 (Effective Challenge Tier 2) clearance ahead of production deployment.
• Delivery Oversight: Direct daily agile standups and sprint planning for the 3-person offshore pod to ensure milestone delivery within the SOW budget.

Role SummaryThe Lead AI Architect & Delivery Manager will serve as the technical authority and engagement lead for the client's TPRM AI Assistant project, backfilling the departing AI Lead. This role is responsible for dri...

• Hands-on Python Expertise: Strong Production Python development, with deep experience in Asyncio, FastAPI, design patterns, and distributed systems.
• Agentic Framework Engineering: 2+ years of direct, hands-on experience building, deploying, and debugging autonomous multi-agent systems in production using LangGraph, AutoGen, CrewAI, or LlamaIndex.
• Function Calling & Structured Outputs: Proven ability to build structured generation pipelines (using Pydantic, Instructor, or native function schemas) to bridge LLMs with traditional software APIs.
• Vector & Relational Storage: Hands-on experience with databases such as PostgreSQL/pgvector, Pinecone, Qdrant, Milvus, or enterprise AI DBs (Oracle 23ai).
• Testing & CI/CD for AI: Experience writing unit tests, integration tests, and regression evaluations for probabilistic, LLM-driven applications.