Headline

AI Engineer at VyasLDM | Generative AI, Agentic AI & LLMs | Retrieval-Augmented Generation (RAG), Python, Machine Learning

About

AI Engineer at VyasLDM, specializing in Generative AI, Agentic AI, and Large Language Models (LLMs). I design and build production Retrieval-Augmented Generation (RAG) pipelines, machine learning systems, and microservice-based AI platforms. At VyasLDM, a legal-technology platform, I contribute to a 6-microservice architecture spanning document processing, LLM-powered legal research, and agentic workflow orchestration. My work includes building RAG pipelines using pgVector and VoyageAI embeddings with LLM-based query-rewriting via Groq, debugging LLM agent routing and classification logic, and designing citation-generation features for legal AI outputs. Previously, I completed a 730-hour interdisciplinary AI/ML research internship at IIMSTC, rated Diamond (top performance tier) on the final assessment. Independently, I have built and deployed end-to-end AI/ML products, including a multi-modal Retrieval-Augmented Generation chatbot with OCR-based document ingestion and a Chroma vector database, and a full-stack agriculture platform published on the Google Play Store. Core competencies: Python, Large Language Models, RAG, Prompt Engineering, Scikit-learn, TensorFlow, FastAPI, PostgreSQL, Vector Databases (Chroma, pgVector), gRPC.

Location

Bengaluru

AI stack

antigravity

claude code

claude

codex

deepseek

gemini

Work experience

Artificial Intelligence Engineer at Vyas LDM, Dec 2025 - Present · 11 mos, On-site

Artificial Intelligence Engineer at Vyas LDM, Dec 2025 - Present · 11 mos, Remote, Contributed to a 6-microservice legal-tech AI platform (Go, Python, gRPC, Restate, PostgreSQL/pgVector), building a production RAG pipeline (pgVector, VoyageAI embeddings, LLM-based query-rewriting via Groq) and durable agentic workflow orchestration for multi-step document ingestion (PDF extraction, classification, annotation). Debugged and fixed LLM agent routing/classification failures, including a query-rewrite misclassification and unbalanced few-shot prompt engineering examples, and designed a same-call citation hover-preview feature generating LLM-based descriptions with no extra call, cache, or database lookup. Diagnosed and fixed production correctness issues, including a billing/credits-deduction bug, a missing database migration risking data loss, and a guardrails bug dropping verification status on LLM-generated legal citations.

Accounting Analyst at IIMSTC (International Institute of Medical Science & Technology Council), Jan 2026 - May 2026 · 5 mos, Interdisciplinary Research Internship (730 hrs) - Bangalore

AI/ML Intern at International Institute of Medical Science and Technology Council, Jan 2026 - May 2026 · 5 mos, Bengaluru, Karnataka, India, Completed a 730-hour interdisciplinary AI/ML research internship, designing, training, and evaluating machine learning models end-to-end for real-world applications, covering data preprocessing, feature engineering, and model validation. Improved model prediction accuracy through iterative training and evaluation cycles using Python and TensorFlow. Rated Diamond (top performance tier, 85%+) on final internship assessment.

Education

Bachelor of Engineering - BE, Artificial Intelligence, HKBK Group of Institutions

Vidyarashmi vidyalaya Savanoor