At Tiger Analytics, I build production backend and AI services with Python, FastAPI, and PostgreSQL. My work includes conversation APIs and a multi-step workflow platform with human approvals, escalation handling, and recovery.
I’ve developed AI applications using RAG, Google ADK, multi-agent workflows, and text-to-SQL. I added query safeguards and AI security controls, and built PromptOps for prompt correction, version management, and evaluation.
I also provision cloud infrastructure and improve deployments across GCP, Azure, and AWS. After moving large structured-data processing outside the model context, I reduced a UI response bottleneck from approximately 65 seconds to under 3 seconds.
