Parthiban Munisamy

AI and data engineer with 5+ years building production-grade AI platforms, with deep experience applying agentic AI and knowledge graphs to industrial and manufacturing domains. Designed a mechanical seal recommendation system backed by a 10,000+ entity engineering knowledge graph, and currently build the graph ingestion layer for an enterprise GenAI platform. Strong across the Azure AI stack, LLM orchestration (LangGraph, MCP, LangChain), classical ML, and large-scale data engineering with PySpark.

Senior Engineering Lead at Persistent Systems (2026-06 – Present)

Market Research GenAI Platform

• Design and build the data ingestion layer feeding a Neo4j knowledge graph — modelling source data into the graph schema of entities and relationships that the platform's GraphRAG retrieval depends on.
• Engineered PySpark and Trino pipelines processing 100M+ records, transforming and loading enterprise data for downstream retrieval and LLM workloads.
• Work alongside the agentic AI layer (Python, LangGraph, GraphRAG), ensuring graph structure and data quality support reliable multi-step reasoning over enterprise knowledge.

Software Engineer at EagleBurgmann KE Private Limited (2023-05 – 2026-06)

Industrial manufacturer of mechanical seals and sealing systems

• Designed an agentic AI system using Python, FastAPI, MCP and LangGraph that interprets engineering requirements and recommends seal configurations, improving recommendation relevance by ~20%.
• Built and maintained a domain knowledge graph of 10,000+ engineering entities in Neo4j and Ontotext GraphDB, modelling product ontologies, materials and application constraints to support semantic reasoning.
• Integrated GenAI and GraphRAG to deliver semantic search across technical product documentation, reducing manual lookup effort by 25% and improving response quality for engineering teams.
• Built a classical ML recommendation system in Python (Pandas, Scikit-learn) using customer interaction and purchase history, improving product suggestion accuracy by ~15%.
• Exposed prediction services via FastAPI to support real-time inference across internal manufacturing applications.
• Developed a form-driven seal selection system in Angular and.NET, streamlining engineering workflows and reducing manual configuration time by ~30%.
• Built interactive dashboards and integrated backend APIs to visualise product data and streamline seal management workflows.

Engineer at L&T Technology Services (2021-01 – 2023-05)

Enterprise Data Integration API

• Developed.NET web applications on Azure App Services and Azure SQL, with Azure DevOps CI/CD pipelines for automated build and release.
• Built a secure.NET Web API on Azure linking backend services to Power Platform, enabling real-time process automation for enterprise users.

M.Tech in Computer Science and Engineering – VIT University (2018 – 2020)

B.E in Computer Science and Engineering – Adhiyamaan College of Engineering (2014 – 2018)