Professional Experience

Leading engineering for AI Teammates, Asana's AI agent product – 23 engineers across 3 teams.

• Took AI Teammates from private beta through general availability.
• Narrowed GA launch scope to focus on reliability and core capabilities; shipped on time.
• Execution reliability improved from 86% to 99%; retention exceeded launch targets.
• Manage AI vendor relationships with Anthropic, OpenAI, and Mistral, including early access programs; the team shipped one of the first production MCP servers.
• Helped reposition AI Retrieval from a single-product feature into a shared platform used across Asana's AI features.
• Context engineering work led to a patent application.

Led AI Retrieval at Asana. Managed 14 direct reports across two teams, including 4 staff engineers.

Highlights

• Production AI: Took 5 AI prototypes to production; the RAG platform grew 35x to 320,000 weekly requests while p95 latency dropped 70%.
• Architecture: Designed a hybrid RAG system (BM25 + dense embeddings with cross-encoder reranking); improved TTFT by 24% and reduced LLM costs by 35%.
• Partnerships: Managed vendor relationships with Anthropic and OpenAI, including early access programs. Featured: Claude 3.5 Sonnet analysis.
• Team: Helped build the AI org structure; supported 8 promotions to Senior and Staff roles; invested in a hiring pipeline for underrepresented candidates.

• Shipped Asana-Jira bidirectional sync, supporting $11M+ in ARR and unblocking 20+ enterprise deals.
• Improved Tableau sync performance from 11 hours to 11 minutes.
• Coordinated scaling work for Asana's largest customer (150k+ users) across 36 engineering teams.

• Joined as an early engineer; became a technical lead and engineering manager as the company grew from Series A to Series C.
• Shipped ML-based recommendations that increased conversion by 25%.
• Rebuilt real-time search filtering with a bit-vector approach, cutting response times from ~7s to 7ms.