I'm Pavan. I'm an AI Engineer, and I like building things that actually run — RAG systems that answer from documents with citations, ML models that catch fraud over streamed events, and the data pipelines underneath that don't break at 2am.
These days I work on generative AI and retrieval (RAG, embeddings, vector search; LangChain and Amazon Bedrock at prototyping level, Azure OpenAI through AI-102) and ML (anomaly detection, forecasting) — standing on batch + streaming data engineering (Spark, Kafka, Delta Lake, Databricks), mostly on AWS. Microsoft Azure AI Engineer Associate (AI-102) and AWS Certified Data Engineer – Associate.
| Project | What it does |
|---|---|
| Cloud Analytics Knowledge Search | RAG knowledge search over 300 synthetic enterprise documents — dense embeddings (MiniLM) + TF-IDF hybrid retrieval, recall@1 87.5%, extractive answers with citations, Streamlit UI |
| Real-Time Retail Intelligence | Kafka → Spark Structured Streaming → Delta Lake platform with ML anomaly detection and forecasting, inline fraud alerts, dbt-tested |
| Retail Sales Analytics | Batch lakehouse (bronze/silver/gold) processing 56K+ orders ($80.96M), 10/10 data-quality checks, PySpark/Databricks port |
| Portfolio | Personal portfolio website — AI Engineer |
- Microsoft Azure AI Engineer Associate (AI-102)
- AWS Certified Data Engineer – Associate
- Databricks Fundamentals Accreditation
