Anmol Srivastava

I’m a Backend & AI engineer who thrives on owning systems end-to-end—from architecture and infra decisions to production deployment—with a strong focus on scalability, reliability, and rapid execution. As the sole backend engineer at Airtribe, I designed distributed, event-driven systems across 5+ business domains, supporting 3K+ MAU and 140+ stakeholders while driving cross-functional delivery.

I also build LLM-powered production pipelines and autonomous workflows: I engineered scalable event-driven RAG and evaluation flows processing 1K+ daily calls, reducing manual review effort by 80% and enabling real-time, evidence-backed coaching. In parallel, I developed stateful agent orchestration for lead lifecycle automation (1K+ active users) using multi-channel notifications and re-engagement logic to improve conversion and reduce no-shows.

Outside of work, I ship research-to-product projects like Sigil (hybrid retrieval with RRF, entity resolution, and performance-tuned memory for coding agents) and Bhunductor (macOS workspace with permission chains, MCP integration, and resilient streaming dedup for resumable context). I’m energized by building retrieval infrastructure and agentic systems that stay dependable under real traffic and fast iteration cycles.