Headline

Data Scientist | Backend & ML Systems | FastAPI, RAG, LLMs, Distributed Systems

About

I build production ML and backend systems that turn complex workflows into measurable business outcomes. I'm a Data Scientist at Mahindra & Mahindra focused on building production-grade ML platforms, distributed backend services, retrieval systems, and intelligent enterprise applications. I've worked across the stack—from PostgreSQL data models and FastAPI APIs to XGBoost/TCN models, RAG pipelines, Azure OpenAI, vector search, Celery/Redis, Docker, and Kubernetes. A few outcomes I'm particularly proud of: 1. Reduced manual effort by 70% through distributed backend automation for enterprise budgeting workflows. 2. Improved retrieval precision by 20% using Azure OpenAI embeddings and Azure AI Search, while cutting manual review turnaround by 50%. 3. Built production systems serving 500+ users and contributed to an estimated ₹7 Cr+ annual business impact through intelligent budgeting recommendations. 4. Built a vehicle-simulation digital twin that saves an estimated ₹4 lakh per simulation by reducing dependence on physical test iterations. What I work on 1. Backend Engineering: Python, FastAPI, REST APIs, WebSockets, Microservices 2. Distributed Systems: Celery, Redis, asynchronous processing, scalable APIs 3. AI/ML: RAG, LLMs, XGBoost, TCN, TensorFlow, NLP, Computer Vision 4. AI Applications: Azure OpenAI, LangChain, LangGraph, vector search, knowledge graphs 5. Data: PostgreSQL, SQLAlchemy, ETL, data modeling, Azure AI Search 6. Cloud & MLOps: Docker, Kubernetes, Jenkins, CI/CD, Azure, GCP I enjoy working at the intersection of software engineering, machine learning, and business impact—especially when the result can be measured in time saved, cost avoided, users served, or revenue generated. If you're building scalable ML platforms, intelligent enterprise applications, or backend systems for AI products, let's connect.

Location

Bengaluru

AI stack

antigravity

claude

github copilot

hugging face

ollama

openai

Work experience

Data Scientist at Mahindra Group, Dec 2025 - Present · 11 mos, Engineered distributed backend pipelines using FastAPI, Celery, and Redis for enterprise budgeting workflows, reducing manual effort by 70% and improving estimation efficiency by 15%. Built a retrieval microservice using Azure OpenAI embeddings and Azure AI Search, improving retrieval precision by 20% and halving manual review turnaround time. Productionized ML inference services and delivered live platform demonstrations to senior leadership, including the President of Mahindra Auto, helping scale the solution to 500+ enterprise users. Led SAST, DAST, and VAPT security validation ahead of production releases, while coordinating Kubernetes deployment across 4 environments: Local, Dev, UAT, and Prod. Contributed across backend architecture, ML deployment, cloud infrastructure, security, and productionization of enterprise AI/ML systems.

Associate Data Scientist at Mahindra Group, May 2025 - Nov 2025 · 7 mos, Architected FastAPI backend services and REST/WebSocket APIs for an enterprise budgeting platform supporting 500+ concurrent users while maintaining high availability under load. Designed the PostgreSQL database architecture, indexing strategy, and migration framework using SQLAlchemy and Alembic from the ground up. Structured data models across 4+ core domains, including vehicle parameters, request history, model registry, and retraining workflows. Built backend foundations that enabled downstream ML, recommendation, retraining, and model-management workflows to operate as a unified production platform.

Education

Bachelor of Technology - BTech, Artificial Intelligence & Data Science, Dr. A.P.J. Abdul Kalam Technical University (AKTU), Lucknow

High School, Mathematics and Computer Science, Don Bosco Academy, Patna