Akhilesh Pal

Designed and deployed LangGraph based LLM agents integrated with LangChain orchestration to generate structured design specifications and production-ready Java code from natural language requirements, reducing manual development effort by 30%. Architected metamodel driven ML code generation framework to automate model implementation workflows, cutting manual engineering time by 25% and standardizing ML integration across projects. Developed a hierarchical multi level classification system using Random Forests, achieving 84.5% accuracy on structured prediction tasks involving layered label dependencies.

R&D Engineer at TCS Research (2023-09 – Present)

• Designed and deployed LangGraph based LLM agents integrated with LangChain orchestration to generate structured design specifications and production-ready Java code from natural language requirements, reducing manual development effort by 30%.
• Architected metamodel driven ML code generation framework to automate model implementation workflows, cutting manual engineering time by 25% and standardizing ML integration across projects.
• Developed a hierarchical multi level classification system using Random Forests, achieving 84.5% accuracy on structured prediction tasks involving layered label dependencies.
• Spearheaded end to end RAG based chatbot pipeline to query business rule documents, enabling SMEs to retrieve accurate, context aware insights and improving decision making efficiency.
• Refactored a Spring Boot Cart microservice, reducing code complexity by 10% and improving overall maintainability.
• Built and optimized Python based Flask REST APIs, applying caching and query optimization to reduce backend latency by 20% and improve microservice scalability.

Research Intern at TCS Research (2023-01 – 2023-06)

• Optimized ML models through systematic hyperparameter tuning and feature engineering, increasing model accuracy from 69% to 75%.
• Constructed an Angular-based UI for dataset visualization, boosting user engagement by 25%.
• Created interactive data visualizations using D3.js to aid researchers in model debugging exploration.

BTech in Computer Science and Engineering – Walchand College of Engineering (2019 – 2023)