Open to opportunities

Rohit Saini

@rohitsaini4

Message

I’m a Lead AI Engineer building GenAI, RAG, and LLM Ops that turn data into measurable business impact.

Zimbabwe

Message

What I'm looking for

I want to build GenAI products end-to-end—RAG/LLM Ops, governance, and measurable business impact—where I can collaborate cross-functionally and turn models into reliable, production-grade systems.

I’m a Lead AI Engineer focused on GenAI that ships: I spearheaded a banking virtual assistant, improving search accuracy to 95% and accelerating customer query resolution using LLM Ops, Python, and prompt engineering. I designed and deployed scalable RAG solutions across 600+ documents, cutting resolution time by 70%, and built validation pipelines generating 7,000+ QA pairs with GenAI (LLM as Judge) plus human-in-the-loop review.

Previously at Zee Entertainment, I led GenAI and recommendation work that moved key metrics—developing production-ready music generation with Meta’s MusicGen, advancing contextual search using vector + knowledge graph methods (including FAISS), and delivering a 9.6% watch-time uplift. Across roles, I also built AutoML-based prediction and forecasting models, implemented AI-driven personalization engines (delivering 80% subscriber conversion lifts), and strengthened model governance/MLOps with feature stores and explainable AI for enterprise compliance.

Experience

Work history, roles, and key accomplishments

Data Science Intern - Real-Time Analytics

Zee Entertainment Ltd.

Jan 2020 - Jun 2020 (5 months)

Developed forecasting and anomaly detection models to enable real-time email alerts, reducing average downtime from 6 hours to 1 hour. Predicted hourly, gateway-level subscriptions from historical and release data and implemented Green/Amber/Red multi-threshold alerting for faster incident response.

Anomaly Detection Predictive Analytics Alerting Incident Response

Data Science Intern - Predictive Maintenance

Siemens Healthineers

May 2019 - Jul 2019 (2 months)

Built a scalable ETL pipeline to process global ultrasound machine logs from ZIP archives and extract structured data without full decompression. Trained an XGBoost classification model for predictive maintenance (83% accuracy) and reduced machine downtime by 10%.

ETL Predictive Maintenance XGBoost Data Structuring Machine Learning

Co-Founder - GharTool

GharTool

Mar 2016 - Feb 2017 (11 months)

Launched a gig marketplace, recruiting 50+ technicians and scaling to 100 DAUs and 20 daily orders through data-driven onboarding. Drove early platform adoption by optimizing onboarding based on user and order signals.

Marketplace Growth Analytics Data On Boarding Performance Metrics Experimentation

Education

Degrees, certifications, and relevant coursework

Indian Institute of Science, Bengaluru

Master of Management, Business Analytics

Earned a Master of Management with a specialization in Business Analytics, covering areas such as deep learning, machine learning, and time series analysis.