At Tezbytes Technologies Pvt. Ltd., I engineered a Python-based data quality engine using FastAPI, LangGraph, and Pydantic. It combines rule-based column checks with schema validation to orchestrate dataset analysis.
I also built a reporting pipeline that brings together rule-based results and LLM processing to generate structured JSON reports on dataset quality, issues, findings, and potential business impact.
For remediation, I developed an LLM-powered workflow that recommends column-specific fixes and SQL, validates syntax and structure, and predicts impact and quality scores. Human-in-the-loop selection and approval gates control execution against Snowflake.
On ProspectIQ, I led project design and coordinated team development of a Python-based B2B sales intelligence prototype. Its research tools and LangGraph workflows combine evidence from web, company, people, and news sources into structured prospect profiles and personalized outreach drafts.
