Aug 2023 – Present · 2 yrs 7 mos

• Architecting the Agnetic AI, Transitive Auth for Agents, Reasoning-based multi-agent platform (MCP-based Agentic AI ecosystem), scaling across Amazon and AWS customers. Designed and published the Remote MCP server with Custom Auth for enterprises and pushed to public aws git labs.
• Architected global employee experience systems aligned with Amazon’s “Earth’s Best Employer” principle.
• Technical leader for AZA (A to Z Assistant), an AI-powered People Platform serving 2.5M+ employees, 60+ countries, 30+ line of businesses.
• Designed diverse agent types (headful/headless, event/schedule-driven) to handle employee workflows, HR, payroll, and beyond.
• Designing the Policy to code infrastructure and Agentic SDLC (Software Development Life Cycle) System in AWS using Gen AI.
• Partnering with AWS to externalize Amazon’s internal AI architecture for enterprise adoption.

Mar 2020 – Aug 2023 · 3 yrs 5 mos

• Designed and delivered the worldwide Time & Attendance system, ensuring accurate pay for employees across 60+ countries.
• Built Amazon’s first unified amazon employee ingress architecture, combining ML, chat, and human support—live globally and integrated into the AtoZ app. Available on Amazon's AtoZ app on Appstore and Google store : https://apps.apple.com/us/app/amazon-a-to-z/id1413909063
• Led design of Fulfillment Operating Service (fOS), an ISA 95–compliant platform standardizing automation/manufacturing across Amazon.
• Granted patent during this time on the synchronous workflow platform design : http://www.pat2pdf.org/patents/pat11210143.pdf
• http://www.pat2pdf.org/pat2pdf/foo.pl?number=11,210,143
• Built warehouse externalization Architecture, leveraging AWS IoT and Machine Learning for anomaly detection for global fulfillment networks.
• Speaker at AWS re:Invent (2019, 2021) presenting large-scale IoT, secure device connectivity, and fulfillment automation architectures. Presenting in 2025.

Jun 2016 – Feb 2020 · 3 yrs 8 mos

• Robotics : Built the communication protocols for Amazon Scout which is currently live at multiple places in US delivering packages to customers. Checkout : https://blog.aboutamazon.com/transportation/meet-scout
• Built the first Chatbot Engine for Amazon using NLP and Machine learning. The chatbot is currently live on both mobile app and desktop of Amazon and most probably you would have also used it (Checkout "Message Us" Chatbot on contact us link of Amazon's mobile app that will resolve your issues)
• https://www.amazon.com/gp/help/customer/contact-us
• Build the algorithm called "Most Relevant Issue" to predict the issue of customer as they show up on the chatbot page with their issue. Checkout this page for seeing it live for you as an amazon customer : https://www.amazon.com/gp/help/customer/contact-us

Oct 2012 – May 2016 · 3 yrs 7 mos

• 1. Built the networking protocol and algorithms to control the communication and physics engine of aut Automated Slam machines we have in Amazon Fulfillment Technology that slams shipping labels on our customer shipment packages (Checkout your amazon's shipping label, looks like random barcodes but believe me its complicated and governs all the best routing algorithms we have built :) )
• 2. Built the backend Customer Package Service and the workflow engine for Amazon which is used by every single packages that is shipped to our customer

May 2010 – Oct 2012 · 2 yrs 5 mos

• 1. Built the Outbound Dock automation for Amazon Fulfillment Technology (AFT) that controlled scheduling of trucks
• 2. Built the Routing automation of Amazon's customer packages that controlled routing of packages on specific chutes of automated Routers