Headline

Machine Learning Engineer | 5 years experience | RAG, Document AI, AI Agents

About

I build ML systems that don't just work in notebooks — they work in production. Fullstack ML Engineer with 4.5 years shipping document AI and production ML systems. I'm that engineer who can take an ML idea from "let's try this" to "it's serving 500 users in production with 99% uptime." My superpower? Speed without sacrificing quality. I shipped an enterprise search system in 2 months that reduced call-center time by 37.5%. Got recognized on the CEO's Wall of Fame for it. I'm also a full-stack engineer turned MLE with more experience building React apps with SpringBoot Java and FastAPI python as backend overlooking deployments via Docker 🔧 WHAT I'VE BUILT Enterprise Search & RAG - Architected RAG search system serving 500+ agents leveraging VertexAI AgentBuilder - Reduced call-center handle time 37.5% (8 mins → 5 mins) - Shipped in 2 months, recognized on CEO's Wall of Fame Document AI Pipelines - Built classification, extraction, clustering systems processing 5 million+ documents daily - Deployed multi-model pipeline (AutoML, CRF, YOLO) with 98% accuracy, 99% uptime - Reduced annotation costs 50% with human-in-the-loop active learning platform improvigng TAT by 800% 💼 CURRENTLY Senior ML Engineer at V-Mart Retail (NSE: VMART) with 550+ stores and 12,000 employees Building RAG systems, LLM-powered workflows, and ML-driven optimization models 💻 TECH STACK ML/AI: Document AI, RAG, NLP, LLMs (Gemini, GPT-4, Claude), Computer Vision Languages: Python, SQL, Java, JavaScript Cloud & MLOps: GCP (Vertex AI, Cloud Run, BigQuery), Docker, Kubernetes, MLflow, Prefect Frameworks: scikit-learn, LangChain, FAISS, TensorFlow, FastAPI, sentence-transformers, ReactJS, NextJS 🎓 BACKGROUND CS Engineering + MBA (Masters' Union) Previously: 3.5 years at Mr. Cooper (PyroAI) building document AI for mortgage industry 📫 OPEN TO ML/AI Engineer roles where I can build production systems that deliver measurable business impact. Especially interested in document-heavy industries (fintech, legal tech, healthcare) and companies that ship fast. Let's connect if you're building something interesting in the ML space.

Location

Bengaluru

AI stack

antigravity

codex

Work experience

Senior Machine Learning Engineer at V-Mart Retail Ltd., Jun 2025 - Present · 1 yr 3 mos, Leading ML and automation initiatives across 500+ retail stores - Built ML-driven interstore transfer optimization using demand forecasting and optimization algorithms; targeting 30% sell-through improvement and 50% logistics cost reduction Tech: Python, scikit-learn, optimization algorithms - Automated Open-to-Buy planning; reduced merchandising cycle from 15 man-days/month to 1 day with daily processing Tech: Python, optimization algorithms, SQL - Built RAG-based knowledge management system using pgvector and Gemini to centralize company-wide operational knowledge Tech: Python, LangChain, PostgreSQL - Developed LLM-powered agentic workflows using custom n8n workflows reducing manual processing from 20-50 hours to <1 hour per workflow; automated document processing and data extraction Tech: n8n, GPT-4, Claude API, Python - Created analytics framework identifying 6% SKU-level revenue leakage across 500+ stores; insights now guide inventory planning Tech: SQL, Python, PowerBI

Strategic AI Initiatives - CEO's Office at V-Mart Retail Ltd., Mar 2025 - Jun 2025 · 4 mos, Conducted strategic analysis and process optimization reporting to CTO - Built analytics framework identifying 6-8% revenue leakage across 500+ stores; defined metrics for inventory planning and assortment optimization Tech: SQL, Python, PowerBI - Led data-driven process audit across 10 departments; prioritized 8 high-impact ML and automation initiatives that became the roadmap for Applied ML function Tech: Process mapping, stakeholder interviews, data analysis - Automated competitor benchmarking pipeline eliminating 100-person manual survey; reduced data collection cycle 50% Tech: Python, web scraping, data pipelines - Delivered centralized analytics dashboards for C-suite; reduced executive decision timelines 50% Tech: PowerBI, SQL, data centralization

Founding Member - Head of Growth & Operations at Bambaii Foods, Jan 2025 - May 2025 · 5 mos, Gurugram, Haryana, India, Making everyday snacking nutritious, flavorful, and guilt-free - Made 6.6 lakhs INR revenue in 6 months - Handled expansion in South India - Partnered with 2 businesses for distribution - Gotten likes from FoodPharmer, Mr.Makhana, SuperYou, Titan Capital and StartupTN for tasting our snacks - Won 1.5 lakhs INR grant for seed round - Handled finance, operations and partnerships

Founding Member - Head of GTM at RugShak, Jun 2024 - Nov 2024 · 6 mos, Gurugram, Haryana, India, Sell funky customisable rugs and carpets made out of New Zealand wool. - Made 3 lakhs INR revenue in 3 months - Made 60K in one day - Handled operations, marketing and tech - Closed 2 B2B deals - Achieved 2.62 ROAS

Machine Learning Engineer 2 at Mr. Cooper, Apr 2023 - May 2024 · 1 yr 2 mos, Remote, Built production ML systems for PyroAI's document AI platform serving enterprise mortgage clients - Architected enterprise search using sentence-transformers + FAISS + hybrid BM25 ranking; served 500+ agents with 25K-100K daily queries, reduced AHT 37.5% (8→5 mins), <200ms p95 latency; shipped in 2 months → CEO's Wall of Fame Tech: Python, sentence-transformers, FAISS, BM25, Elasticsearch, FastAPI, GCP My role: Technical lead, architecture design, end-to-end implementation - Built HITL annotation platform with 2 MLEs using active learning; improved annotation throughput 8x (8 hours→1 hour), reduced labor costs 50% (68→34 annotators), integrated with model training pipelines Tech: Python, FastAPI, PostgreSQL, active learning, annotation workflows My role: Led development team, designed architecture, built backend APIs - Deployed multi-model document AI pipeline processing 5M+ docs/day (3000 pages/min) with 98% accuracy and 99% uptime; used AutoML for classification, CRF for NER, YOLO/CenterNet for extraction Tech: Python, GCP AutoML, CRF, YOLO, CenterNet, TensorFlow, Docker, Kubernetes My role: Pipeline architecture, model integration, production deployment, monitoring - Re-architected ML model serving from GCP AI Platform to Kubernetes; maintained throughput while reducing infrastructure costs 50% with better scaling control Tech: Docker, Kubernetes, GCP Cloud Run, Cloud Build, MLflow, Python My role: Infrastructure migration, containerization, deployment optimization - Built RAG prototype using Vertex AI and Gemini; served as technical SPOC for Google Cloud Consulting on GenAI initiatives for call-center automation Tech: Vertex AI, Gemini, LangChain, RAG architecture, prompt engineering My role: Prototype development, Google Cloud liaison, LLM evaluation

Software Engineer 1 (Machine Learning) at Mr. Cooper, May 2022 - Mar 2023 · 11 mos, Developed ML infrastructure and document AI models for PyroAI platform - Re-architected ML serving infrastructure using Docker containerization; migrated from monolithic deployment to microservices reducing inference costs 50% and improving developer productivity 30% Tech: Docker, Kubernetes, GCP Cloud Run, Python, model serving optimization My role: Infrastructure redesign, containerization, deployment pipeline setup - Built unsupervised document clustering pipeline using K-means; automated classification reducing manual work 50%, 99% uptime; shipped microservice API in 2 weeks Tech: Python, scikit-learn, K-means, TF-IDF, Flask/FastAPI, production deployment My role: Model development, API implementation, end-to-end delivery - Built MLOps pipeline for CRF (Conditional Random Fields) model deployment; automated training and deployment workflows enabling faster releases Tech: Python, CRF, MLflow, CI/CD pipelines, automation scripts My role: Pipeline development, deployment automation - Developed checkbox/radiobutton extraction using computer vision + regex; improved overall document extraction accuracy by 20% Tech: OpenCV, Python, image processing, regex pattern matching My role: CV algorithm implementation, accuracy optimization - Built REST API for model deployment to monitor and deploy GCP AI Platform models and Vertex AI pipelines; bridged developer-DevOps gap for faster iterations Tech: Python, Flask/FastAPI, GCP AI Platform, Vertex AI APIs My role: API development, DevOps integration - Led early development of HITL annotation feedback loop; built ML pipelines and APIs for model-annotator interaction workflows Tech: Python, annotation pipelines, feedback loop architecture My role: Led 2 MLEs, designed initial architecture, built APIs

Software Engineer - ML at Mr. Cooper, Jul 2021 - May 2022 · 11 mos, Built document processing infrastructure and ML pipelines for mortgage automation - Developed distributed thumbnail generation service using GCP Cloud Functions + Pub/Sub; processed 1M+ PDF images simultaneously generating multiple thumbnail sizes with 98% success rate Tech: Python, GCP Cloud Functions, Pub/Sub, Cloud Storage, distributed processing My role: Service architecture, implementation, production deployment - Optimized PDF bursting microservice handling 5M+ documents daily; fine-tuned resource allocation and monitoring reducing p95 latency by 98% Tech: Python, microservices, resource optimization, performance monitoring My role: Performance analysis, resource tuning, optimization - Built document extraction workflows and ML infrastructure enabling PyroAI platform to process millions of mortgage documents Tech: Python, OCR pipelines, document processing, ML infrastructure My role: Pipeline development, workflow implementation - Implemented search, sort, and filter capabilities in document management UI; improved user experience for document retrieval workflows Tech: React, JavaScript, UI development, search optimization My role: Frontend implementation, UX improvements

Content Reviewer at Section EngEd Program, Oct 2020 - Apr 2022 · 1 yr 7 mos, 💼 Role - Write technical tutorials as a contributor - As a peer-reviewer, review tutorial ideas and approve them - Verify basic checklists while submitting a PR - Verify for grammar and plagiarism check of tutorials - Verify and test code snippets in the tutorials - Approve tutorials for publishing ⚔️ Key Accomplishments - #1 active contributor with around 690+ commits editing 35K+ lines in the repo - Written 13 technical tutorials with over 100K+ views - Reviewed 200+ tutorials and approved them for publishing - Reviewed 100+ tutorial ideas before allowing contributors to work on them - Was part of a discussion panel hosted by the Community Manager on "Boosting your Software Career Journey". You can view the post here - (https://www.linkedin.com/posts/section-enged-program_enged-stories-how-dev-communities-can-help-activity-6958842014868148226-49iE?utm_source=share&utm_medium=member_desktop)

Education

Master of Business Administration - MBA, Technology and Business Management, Masters' Union

Bachelor of Technology, Computer Science and Engineering, Amrita Vishwa Vidyapeetham

Public links

github: https://github.com/srishilesh