Adhi Shakthi

AI Engineer and Retrieval Engineer with experience building production-grade Retrieval-Augmented Generation (RAG) systems, retrieval pipelines, and AI infrastructure. Passionate about Information Retrieval, Large Language Models, semantic search, observability, and benchmarking.

Built

PilotMaster (pilot-master.vercel.app) , a modular retrieval intelligence platform featuring document intelligence, execution observability, and benchmarking capabilities for evaluating and optimizing RAG systems. Previously worked on OCR document processing and machine learning pipelines, with a strong interest in designing scalable AI systems that move beyond simple chatbot applications.

OCR Engineer at First American India (2023-09 – 2024-08)

Developed complex regex-driven extraction and validation pipelines for processing noisy OCR-generated American real estate documents, including normalization of inconsistent textual patterns, field-level data cleaning, and rule-based parsing of semi-structured property records to improve extraction accuracy and downstream document reliability using Tesseract.

Intern at Defence Research and Development Laboratory (2022-02 – 2022-03)

During this internship, developed a two-dimensional finite-volume grid preprocessing utility in C for converting structured computational grids into CFD solver-compatible data structure formats. The generated data structures were thoroughly verified through both manual and numerical validation techniques to ensure preprocessing accuracy and consistency.

MS – Scaler Academy x Woolf College - Neovarsity (2026)

Specialisation in Software Development – CDAC Hyderabad (2023)

BE/B.Tech/Bs – MVSR Engineering College (2022)