Skip to content

RESUME.md

Latest commit

 

History

History
23 lines (15 loc) · 1.46 KB

RESUME.md

File metadata and controls

23 lines (15 loc) · 1.46 KB

Resume Bullets

Project Title

AI SQL Analyst | Production-Style Text-to-SQL Analytics Platform

Short Resume Version

  • Built a production-style text-to-SQL analytics app using FastAPI, PostgreSQL, Docker, Kubernetes, and Terraform, enabling natural-language questions over structured SaaS warehouse data.
  • Implemented SQL guardrails that enforce read-only queries, block unsafe statements and unknown tables, add row limits, and inject workspace_id filters for multi-tenant query isolation.
  • Added API key authentication, query telemetry, history, latency metrics, chart metadata, and an analyst console for inspecting generated SQL and results.
  • Created automated text-to-SQL eval suite and CI pipeline with unit/API tests plus PostgreSQL integration testing through GitHub Actions.

Interview Expansion

  • Designed the system so LLM output is treated as untrusted code: generated SQL must pass validation before execution.
  • Supported both SQLite fallback for local development and PostgreSQL for production-style deployment.
  • Added Kubernetes manifests with Deployment, StatefulSet, Services, ConfigMap, Secret, readiness/liveness probes, and HPA.
  • Added Terraform example for provisioning the Kubernetes app resources into an existing cluster.

One-Line Portfolio Description

Production-style AI analytics platform that converts business questions into guarded, workspace-scoped SQL with FastAPI, PostgreSQL, Docker, Kubernetes, Terraform, CI, and automated evals.