Senior AI Software Engineer with over 8 years of experience in software development and 2+ years of experience developing AI-driven tools and solutions. Expertise in LangChain, LangGraph, and RAG with hands-on development of multi-agent applications, Agentic workflows, and knowledge graphs systems at scale.
LangChainLangGraphRAG SystemsMulti-Agent SystemsNeo4jLLMsPrompt EngineeringFastAPILangServeGradioStreamlitMLflowHugging Face
Became an AI advocate within the team and organization. Led training sessions for Developers on integrating and using AI.
Always researched and created POCs about the latest advancements in LLMs, fine-tuning techniques, and AI tools and shared findings with the teams as new advancements come up.
Led the development of multi-agent applications using LangChain and LangGraph frameworks for efficient LLM and RAG-based systems.
Collaborated with teams to develop proof-of-concept projects exploring RAG solutions, including Hybrid RAG, GrpahRag and Agentic RAG.
Developed CI/CD pipelines for testing, and deployment of the agents.
Integrated knowledge graphs using Neo4j, improving data processing and retrieval.
Demonstrated expertise in a broad spectrum of AI tools including ChatGPT, co-pilot, and Cody, RelicX applying advanced prompt engineering techniques to maximize tool effectiveness and productivity.
Developed an instruction-tuned dataset for the purpose of fine-tuning test automation LLMs in a local and secure manner.
Fine-tuned LLaMA3 for specific use cases and domain-specific tasks.
Conceptualized and executed multiple proof-of-concept projects exploring the application of various AI tools in our project.
Developed local AI tools fine-tuned for the specific use cases for the applications for increasing productivity.
Fine-tuned and quantized different LLMS to enhance model performance in local environments, reducing resource consumption.
Tested our chatbot with different prompt injections and jail breaking techniques.
Designed and implemented custom data preprocessing and augmentation pipelines to improve model performance.
Used MCP tools to create a ticket to PR workflows in Cursor, to release features really really fast.
Created MCP tools to automate workflows without creating agents.
