Lokesh Reddy Dodla

I'm Lokesh Reddy Dodla, an AI engineer and full-stack developer based in Hyderabad. I build AI products end to end: the agent or RAG layer, the FastAPI/Django backend, the React front end, and the deployment that keeps it running. I currently hold two roles. At OneSig (since July 2026 as part-time) I work on a UAE e-invoicing readiness platform: spreadsheet-to-schema mapping, validation against the PINT AE rule set, and the workflow, milestone and audit models behind it, built with Django/DRF, React + TypeScript and AWS. At AiHealthClinic (since August 2025 as full-time) I build healthcare AI services in production, including a patient-scoped RAG platform over large medical records, a coaching agent that reads InBody scans, rule-based lab-report classification and appointment-booking APIs over PostgreSQL and MongoDB. Outside work I ship things I can point to. UAE Invoice Check (uaeinvoicecheck.web.app) is a free tool that runs all 303 official PINT AE rules against a business's invoice export and explains what will fail in plain language; it runs on Google Cloud Run and the code is public on GitHub. I also built a tool-calling operations agent on the Anthropic and Groq APIs where deterministic code decides the facts, the model decides the wording, and nothing is sent without a human approving it, with an eval suite to prove it. How I work: I prefer small, testable pieces over clever ones, I measure before I claim, and I'd rather deploy an honest version today than a perfect one later. Stack: Python, FastAPI, Django, TypeScript, React, PostgreSQL, MongoDB, LLM tool use and RAG, Docker, Google Cloud, AWS, Vercel. Portfolio: lokeshreddydodla.me · GitHub: github.com/Lokesh1028 · LinkedIn: linkedin.com/in/lokesh-reddy-dodla