Senior Data Engineer | Building Gen AI Applications
Building RAG systems, LLM agents, and AI-powered data tools.
Text-to-SQL system with production monitoring
Problem: Business users need to query data without writing SQL
Solution: Natural language interface powered by GPT-4 with LangGraph orchestration
Tech Stack: Azure OpenAI GPT-4 β’ LangGraph β’ DuckDB β’ Prometheus β’ Grafana
Key Features:
- Natural language to SQL conversion
- CSV/Excel file analysis capability
- Real-time performance monitoring
- Complete observability stack with metrics
Hands-on implementations exploring LLM capabilities
Experiments:
- RAG Systems: ePub, PowerPoint, PDF, Word document processing with FAISS
- LLM Output Structuring: Using Pydantic for structured responses
- Function Calling: Dynamic weather chatbot with tool use
- Context-Aware Chatbot: E-commerce assistant with memory and chains
Tech Stack: FAISS β’ LangChain β’ Azure OpenAI β’ Pydantic β’ Vector Embeddings
Focus: RAG patterns, structured outputs, function calling, memory management, chunking strategies
LangGraph workflow orchestration
Problem: Complex agent coordination and state management
Solution: Multi-agent workflow system using LangGraph
Tech Stack: LangGraph β’ Azure OpenAI β’ MySQL β’ RAG Integration
Key Features:
- Multi-agent communication patterns (booking, housekeeping, customer service)
- Shared state management across agents
- RAG-enhanced response generation
- Complete workflow orchestration with Streamlit UI
Multi-agent SQL generation and validation
Problem: Secure SQL generation with validation
Solution: Multi-agent system using CrewAI framework
Tech Stack: CrewAI β’ Azure OpenAI β’ Multi-agent orchestration
Key Features:
- SQL generation agent
- SQL validation agent
- Security compliance checking
- Cost monitoring
Generative AI: LangChain β’ LangGraph β’ CrewAI β’ Azure OpenAI β’ FAISS β’ Pydantic
Data Engineering: Python β’ PySpark β’ DuckDB β’ SQL
Monitoring & DevOps: Prometheus β’ Grafana β’ Docker β’ Terraform
Databases: MySQL β’ Vector Databases
- Databricks Generative AI Engineer Associate (2025)
π View Credential - Databricks Certified Data Engineer Associate (2025)
π View Credential
Learning projects and proof-of-concepts exploring generative AI applications. For my professional background, visit my LinkedIn profile.