Prathap E

AI Engineer and LLM Developer with 1.5+ years of production experience building agentic AI systems, RAG pipelines, multi-agent orchestration workflows, and LLM-powered enterprise applications. Achieved 93% accuracy in fine-tuned BERT transformer models, architected RAG pipelines with LangChain and vector databases. Experienced in Python, Azure AI stack, FastAPI, Docker, CI/CD, and MLOps evaluation frameworks.

Adept at translating complex business problems into scalable AI solutions across agile, cross-functional teams.

AI Data Engineer - Klodev - Coimbatore, India

(2025-08 - 2025-11)

Full-time | ETL, Data Pipelines, MLOps & Enterprise Data Engineering

• Constructed LLM-enhanced ETL pipelines processing 100M+ records monthly with automated entity extraction, metadata enrichment, CDC, and SCD2 tracking on Azure Databricks.
• Engineered reusable Python library standardising data cleaning across 20+ enterprise pipelines, reducing manual effort by 60%; deployed as Databricks artifact for 12 production pipelines.
• Accelerated Databricks query performance by 40% through Z-Ordering, Delta Lake compaction, partition optimisation, broadcast joins, and cluster auto-scaling on 10TB+ datasets.
• Implemented CI/CD pipelines via Azure DevOps for automated deployment and testing; administered Unity Catalog with role-based access controls and data lineage tracking for 100+ assets.

AI Data Engineer Intern - Klodev - Coimbatore, India

(2024-07 - 2025-07)

Internship | AI Engineering, LLM & Generative Intelligence

• Fine-tuned BERT transformer models achieving 93% F1-score in NER address parsing across 60M+ global records; deployed production Flask API processing 5,000+ queries/day with sub-second latency.
• Architected end-to-end RAG pipeline using LangChain, vector databases, and hybrid semantic search achieving 69% precision in demographic extraction for 10,000+ companies, improving data completeness by 35%.
• Developed custom MCP server connecting Outlook to Claude AI via Microsoft Graph API, enabling autonomous agentic email operations for 500+ daily messages hands-free.
• Designed multi-agent orchestration workflows with prompt chaining, tool calling, and memory management using LangChain and agent frameworks for enterprise automation

B.Tech - Artificial Intelligence & Machine Learning - Sri Shakthi Institute of Engineering and Technology (2021-10 - 2025-05)