Engineering Lead, AI & Data Platform, Company confidential (CPA & advisory firm, ~700 people)
Mar 2026 - PresentLead a platform team building the firm's in-house production AI and data platform.
- Own the canonical AI developer platform: container-based, API-driven, on ARM compute with eBPF networking, running AI agents, model services, and workflow automation in production across multiple engineering teams.
- Standardized model inference on a unified serving runtime behind a gateway layer with routing, fallbacks, A/B experimentation, and inference-time optimization.
- Delivered enterprise retrieval pipelines combining vector and graph search, powering a sovereign research-intelligence layer with full data lineage on a lakehouse stack.
- Designed streaming agent orchestration: model tool calls dispatched to durable workflows over pub/sub messaging with live client progress.
- Built reusable platform primitives for authentication and fine-grained, relationship-based authorization; hardened with centralized secrets management and admission policies.
- Ran a DORA baseline across 180 repositories and 46 projects; produced the build-vs-buy economic justification adopted at firm leadership level.