Jan 2024 – Jan 2026 · 2 yrs · Bengaluru, Karnataka, India · On-site

• 𝐁𝐮𝐲𝐢𝐧𝐠 𝐜𝐚𝐫 𝐩𝐚𝐫𝐭𝐬 𝐨𝐧𝐥𝐢𝐧𝐞 𝐬𝐡𝐨𝐮𝐥𝐝𝐧’𝐭 𝐟𝐞𝐞𝐥 𝐥𝐢𝐤𝐞 𝐚 𝐠𝐚𝐦𝐛𝐥𝐞.
• I designed and built AI/LLM-driven fitment systems at Amazon that match any automotive part or accessory to the right vehicle. From filters to tires to complex components, my large langauge models eliminate guesswork and give shoppers confidence before they click “Buy.”
• 𝐖𝐢𝐥𝐥 𝐭𝐡𝐞𝐬𝐞 𝐭𝐰𝐨 𝐩𝐫𝐨𝐝𝐮𝐜𝐭𝐬 𝐰𝐨𝐫𝐤 𝐭𝐨𝐠𝐞𝐭𝐡𝐞𝐫?
• I turned that question into an AI challenge. I architected and launched an AI/LLM-powered compatibility engine across Amazon’s massive catalog — spanning electronics, accessories, home goods and beyond — so customers instantly know what fits with what, reducing uncertainty and returns at scale.

Sep 2022 – Dec 2023 · 1 yr 3 mos · Bengaluru, Karnataka, India · On-site

• 𝐄𝐯𝐞𝐫 𝐰𝐨𝐧𝐝𝐞𝐫𝐞𝐝 𝐡𝐨𝐰 𝐲𝐨𝐮 𝐢𝐧𝐬𝐭𝐚𝐧𝐭𝐥𝐲 𝐬𝐞𝐞 𝐭𝐡𝐞 𝐫𝐢𝐠𝐡𝐭 𝐬𝐢𝐳𝐞 𝐢𝐧𝐟𝐨 𝐨𝐧 𝐀𝐦𝐚𝐳𝐨𝐧?
• I built and scaled the AI/LLM pipelines that extract, standardize and surface accurate dimension data across millions of products — from oversized furniture to tiny accessories — transforming a messy sea of supplier specs into a clean “Will it fit?” experience for shoppers worldwide.

Jun 2020 – Sep 2022 · 2 yrs 3 mos · Chennai, Tamil Nadu, India

• 𝐌𝐢𝐠𝐫𝐚𝐭𝐚 – 𝐈𝐧𝐭𝐞𝐥𝐥𝐢𝐠𝐞𝐧𝐭 𝐜𝐥𝐨𝐮𝐝 𝐦𝐢𝐠𝐫𝐚𝐭𝐢𝐨𝐧 𝐚𝐭 𝐬𝐜𝐚𝐥𝐞: I architected and delivered Migrata, TCS IP-safe a highly scalable migration platform that not only moves enterprises from any on-premise database to GCP with minimal effort, but also embeds ML models to transform raw migration data into actionable insights during the process. Launched during the 2019-22 cloud-shift wave, 𝑴𝒊𝒈𝒓𝒂𝒕𝒂 𝒊𝒔 𝒔𝒕𝒊𝒍𝒍 𝒊𝒏 𝒖𝒔𝒆 𝒕𝒐𝒅𝒂𝒚 𝒂𝒏𝒅 𝒓𝒆𝒎𝒆𝒎𝒃𝒆𝒓𝒆𝒅 𝒃𝒚 𝒇𝒐𝒓𝒎𝒆𝒓 𝒄𝒐𝒍𝒍𝒆𝒂𝒈𝒖𝒆𝒔 𝒂𝒔 𝒂 𝒈𝒂𝒎𝒆-𝒄𝒉𝒂𝒏𝒈𝒆𝒓. It has been sold as a platform-as-a-service across multiple domains, cutting the time and cost of cloud migrations.

Feb 2019 – Jun 2020 · 1 yr 4 mos · Chennai, Tamil Nadu, India

• 𝐢𝐠𝐧𝐢𝐨 – 𝐀𝐮𝐭𝐨-𝐌𝐨𝐝𝐞𝐥𝐥𝐢𝐧𝐠: I have contributed in scale to 𝐢𝐠𝐧𝐢𝐨, a product with built-in auto-modelling and hyper-parameter tuning that accelerates experimentation for data scientists across the organisation. It’s been adopted widely inside TCS and sold via the TCS Marketplace to external clients.
• 𝐃𝐫𝐢𝐯𝐢𝐧𝐠 𝐌𝐋 𝐢𝐧𝐭𝐨 𝐞𝐧𝐭𝐞𝐫𝐩𝐫𝐢𝐬𝐞 𝐬𝐮𝐩𝐩𝐨𝐫𝐭 𝐰𝐨𝐫𝐤𝐟𝐥𝐨𝐰𝐬: : As part of the Machine First Delivery Model (MFDM) Centre of Excellence, I built ML systems to understand the intent of support tickets, automatically classify them, and route to automation, semi-automation or manual resolution — reducing turnaround time and improving accuracy for multiple TCS customers.