anjali ega

Anjali Ega

Software engineer, five years in production systems. Most of that was a multi-year FedEx engagement through Wipro — Spring Boot services over JMS for four field applications. The current focus is AI applications: retrieval, agents, orchestration, and the engineering work that keeps them shippable instead of demo-only.

Work authorized in the US (H4 EAD). Open to senior engineering roles. Dallas, TX.

Currently building

Lumen — Natural-language analytics with a YAML semantic layer and schema-aware SQL validation that catches hallucinated columns before execution.

Projects

  • Atlas — Enterprise search with permission-aware retrieval across Slack, Notion, Drive, GitHub.
  • Sentinel — GitHub app for AI code review with repo-aware context.
  • Quorum — Multi-agent orchestration with typed message contracts and cost ceilings.
  • Echo — Real-time voice AI for sales calls, sub-300ms transcription.

See all projects

Writing

Schema-aware validation: how to keep an LLM from hallucinating SQL

All writing