Work Experience
CD
Community Dreams Foundation
AI Engineer
Feb 2026 – Present
USA
Currently Here85%Grounded Accuracy
20%Less Manual Handling
50+Documents Indexed
- ▸Built a RAG-based AI application using LangGraph, the OpenAI API, and an open-source vector store (FAISS), indexing 50+ internal program documents and achieving ~85% grounded-response accuracy across 30 evaluation questions.
- ▸Developed LangGraph-based agentic workflows with 3 tool integrations, enabling agents to retrieve context, execute functions, and complete multi-step requests while reducing manual handling by ~20%.
- ▸Integrated the AI assistant with a lightweight SQLite database of program records, enabling conversational access to curated organizational data and reducing typical lookup time from ~6 minutes to under 2 minutes.
LangGraph
OpenAI API
FAISS
Pydantic
SQLite
Zapier
Ac
Accenture
Data Engineer
May 2021 – Oct 2023
India
1M+Records / Run
25%Fewer QA Failures
35%Less Manual Work
- ▸Built ingestion, transformation, and migration pipelines using Azure Data Factory, Azure Data Lake Storage, PySpark, and SQL, processing 1M+ records per pipeline run across structured and semi-structured sources.
- ▸Implemented Bronze, Silver, and Gold data layers using Medallion architecture, reducing failed data-quality checks by ~25% and improving availability of curated datasets for downstream analytics.
- ▸Developed Python-based data services and FastAPI REST APIs, reducing recurring manual data-extraction requests by ~35% and enabling self-service access for analytics applications.
Azure Data Factory
PySpark
SQL
Medallion Architecture
FastAPI
Azure OpenAI