RAHUL KHANDEKAR

Projects:

1. Supply Chain.

Contract Analysis -

• Developed a GenAI-powered tool to analyze contracts for legal soundness and future risk, enhancing accuracy and sustainability using Azure OpenAI and LangChain.

Demand Sensing -

• Built a solution to detect real-time external factors (e.g., social media, politics) affecting demand forecasts, using sentiment analysis and trend detection.
• Role and Responsibility (Gen-AI Developer):
• Designed and implemented an intelligent chatbot using Python.
• Integrated Azure OpenAI models for natural language understanding and generation.
• Enabled chatbot support for contract analysis and demand forecasting.
• Built interactive user interfaces using Streamlit.
• Performed sentiment analysis and trend detection from external sources (e.g., social media, political events).
• Applied prompt engineering and fine-tuning to optimize model performance.
• Technologies Used: Python, Streamlit, RAG, Azure OpenAI (GPT models), and LangChain.

2. Intelligent Pipeline.

• Developed a metadata-driven ETL pipeline using the medallion architecture (Bronze-Silver-Gold) to automate data ingestion, transformation, and curation.
• Key Responsibilities:
• Designed and implemented a multi-layered ETL pipeline using Python and SQL Server.
• Built robust data validation and rejected record-handling logic to ensure data integrity.
• Developed modular Python components for metadata extraction and configuration management.
• Implemented automated flagging, auditing, and reconciliation for pipeline monitoring and governance.
• Technologies Used: Microsoft SQL Server, Python, Azure OpenAI, Pandas, and Excel.

3. Data Quality Report Generation.

• Created an automated data quality validation framework using Azure OpenAI to dynamically generate Python validation logic from business rules.
• Key Responsibilities:
• Developed a Python-based engine for rule-based validation of tabular datasets.
• Integrated Azure OpenAI to auto-generate and execute validation functions.
• Implemented fuzzy logic for column-rule matching, and robust error handling.
• Built modular components for rule ingestion, validation execution, and reporting.
• Collaborated with stakeholders to translate business rules into executable logic.
• Technologies Used: Python, Pandas, Azure OpenAI, and Streamlit.