Headline

AI Product Manager at Unilever | AI Products, Last-Mile Ops, B2B/B2C Commerce | 0→1 to Scale | SPJIMR | PEC

About

I build products that work at scale, not just at launch. Over the last 7 years, including 4 years as Product Manager at Hindustan Unilever, I have shipped AI-powered and consumer-facing products used by 2.5M+ outlets, retailers, and delivery partners across 21+ countries. My work sits at the intersection of AI/ML personalisation, ops-tech, and B2B commerce. At Hindustan Unilever, I designed iQ Nexus, a neural-network personalisation engine that has delivered 150M+ SKU recommendations globally across 2.5M+ outlets and drives meaningful incremental revenue moving the business from static, tree based, rule-based logic to models that capture the patterns traditional approaches could not. I work closely with data science across the full product loop: defining model inputs, shaping evaluation criteria, measuring success in business terms, and driving adoption in the field. Alongside this, I own the product vision for a global Distributor Management Platform built from 0 to 1 across multiple 8+ SA/SEA markets, and have built last-mile delivery and warehouse apps at quick-commerce scale. What I enjoy most is the unglamorous part of AI products - solving cold-start for new outlets, making recommendations explainable enough for a salesperson to act on, and driving the change management that turns a model into daily usage. Outside work, football is my other obsession. I analyse it the same way I analyse products: patterns, systems, and why things break down under pressure. Dream big. Work hard. Stay focused. Be surrounded by good people.

Location

Mumbai

AI stack

antigravity

codex

Work experience

AI Product Manager at Unilever, Sep 2023 - Present · 3 yrs, Bengaluru, Karnataka, India, • AI Product - Recommendation Engine : Built and productionised a neural-network recommendation engine delivering 150M+ AI-driven product recommendations across 2.5M+ outlets in 21 countries. Owned the full AI product lifecycle from defining model inputs and evaluation criteria to measuring business outcomes in revenue terms along with model accuracy. Stack: Azure Data Lake, collaborative filtering, RFM signals, seasonality modelling. Drove 5% to 8% uplift in sales hit rate. • AI Product - Demand Prediction: Extended the recommendation engine with an ML-driven demand prediction layer that forecasts the right quantity of each product a retail outlet needs monthly, replacing static formula-based logic with a data-driven framework using purchase history patterns, seasonality signals, and demand indicators. Improved prediction accuracy by moving 16%+ more outlets into the exact demand bucket, reduced under-stocking risk by 39%, and increased recommendation value coverage from 35% to 74% of actual sales. • B2B Commerce Platform & Last Mile Delivery - Owned the product vision and strategy for a global B2B Commerce platform (Distributor Management System) handling the end-to-end Order to Fulfilment cycle for distributors across South Asia and Southeast Asia, built on a scalable GCP and Azure micro-service architecture, live in 8+ SA/SEA markets integrated with logistics, warehouse, and field sales systems.

Unilever Future Leaders Programme at Unilever, Jun 2022 - Aug 2023 · 1 yr 3 mos, • Owned market expansion and growth roadmap for Ludhiana, partnering with distributor teams to strengthen retail footprint and network coverage; drove 110%+ target achievement for distributor teams and key account executives; defined a 3-year roadmap to 2x GMV through distributor enablement, targeted network expansion, and data-driven outlet prioritisation. • Led cross-market supply chain benchmarking across 4 countries, assessing end-to-end logistics costs, depot operations, and last-mile distribution efficiency; translated benchmarking insights into structural cost-saving strategies and network optimisation recommendations adopted across markets.

Unilever Leadership Internship Programme at Unilever, Sep 2021 - Oct 2021 · 2 mos, • Launched and scaled a production-grade AI voice ordering assistant on the Shikhar B2B app, handling real transactional intent (order placement) in local languages for non-digital-native users. Drove significant MAU growth, feature activation, and order frequency improvements through structured funnel drop-off analysis and A/B testing, closing the loop between product and data science from feature specification to business KPI measurement. • Designed and productionised a next-generation low-code omnichannel chatbot platform enabling multi-channel conversational AI deployment across WhatsApp, app, and web without per-channel engineering effort, reducing time to deploy new conversation flows from weeks to days. • Owned product delivery of a Virtual Customer Support AI assistant for B2B e-commerce, handling inbound support queries through NLP-based intent classification and automated resolution workflows, reducing reliance on human agents for tier-1 queries.

Full Stack Product Developer at Egregore Labs - Robana & Romulus, May 2019 - Jul 2020 · 1 yr 3 mos, Delhi, India, • Owned end-to-end build of an equity research and investment analytics platform for a global investment bank delivering structured data pipelines, research workflows, and analytics dashboards with 100% on-time launches across all modules • Built a Data Enrichment and intelligent T&E expense auditing system using automated anomaly detection and policy-violation flagging to replace manual audit cycles; achieved 80% reduction in audit costs • Designed and productionised an end-to-end document processing pipeline using automation to extract, classify, and route financial documents, eliminating manual handling at scale • Built an integrated financial research collaboration platform that unified fragmented data sources and research workflows into a single system eliminated information silos by 70% across analyst teams • Architected a modular data access layer decoupling business logic from data retrieval, enabling reusable, scalable components across the platform

Software Product Developer ll at XLPAT, Jul 2018 - Apr 2019 · 10 mos, Sahibzada Ajit Singh Nagar, Punjab, India, • Built the world's first automated commercialised novelty checker - an AI-powered prior art search system scanning 10 crore+ patent records and 20 lakh scholarly works in under 2 minutes, replacing weeks of manual attorney research with automated ML-driven eligibility assessment • Owned an IP monetisation module that used semantic similarity models to surface licensing and commercialisation opportunities from a corpus of 10 crore+ patents contributed approximately 30% of business revenue • Developed a blockchain-based IP solution for IPBC Global 2018, applying distributed ledger technology to patent provenance and ownership verification

Software Product Developer at XLPAT, Jun 2017 - Jun 2018 · 1 yr 1 mo, Mohali, Punjab, India, • Built a semantic search and competitive intelligence engine for patent discovery using NLP-based similarity matching, achieving 91% precision enabling IP analysts to surface prior art and competitive filings in seconds rather than days • Led migration from LAMP to MEAN stack to support real-time search and analytics at scale across a corpus of tens of crore patent records

Patent Analyst at TT Consultants, Jun 2017 - Jun 2017 · 1 mo, Mohali, Punjab, India, • Executed various kinds of projects including patentability, Infringement • Received Appreciation for Intel Patentability Search

Education

PGDM, Information Management, SPJIMR SP Jain Institute of Management & Research

Bachelor of Engineering - BE, Electronics & Electrical Communication Engineering, Punjab Engineering College