Dec 2021 – Jun 2022 · 6 mos

• Bit of Europe, mostly South America

Mar 2019 – Nov 2021 · 2 yrs 8 mos · London, United Kingdom · On-site

• Trust & Safety
• A YouTube talk on my work here - https://www.youtube.com/watch?v=hdlW2kKkm3A
• Built Neural Networks, GBDT based models to catch compromised accounts & spammers. Decreased account hacking & spam activity by >30% in key markets. - Handled the entire lifecycle of ML model development.
• Integrated A/B testing framework into ML models to be able to compare multiple ML models, thereby making data-driven decisions for any change to the ML models. ~50% reduction in the number of ML models.
• Build rule-based detection and mitigation to catch abuse types like two-factor pin set by a hacker, account locked due to SMS limit exhaust, leading to ~80% reduction in the customer support ticket.
• Implemented ‘Logout’ enforcement for compromised accounts instead of banning them, led to a decrease in the customer support tickets by ~70%, increased user retention & improved user experience.
• Helped the team grow from <10 Engineers to 25+ Engineers.

Jul 2017 – Jan 2019 · 1 yr 6 mos · Bengaluru Area, India

• Nutanix ships private cloud for the enterprise. We had an in-house NoSQL (key-value) database for storing entities. It is a cache written in C++ using Cassandra for persistence.
• The cache supported ACID operation, search, time-series queries, and efficient state change notifications.
• I built some features for the database, a client library in Python to access the database and made the database run on Kubernetes.

Oct 2015 – Jun 2017 · 1 yr 8 mos · Bengaluru Area, India

• My work at Swiggy remains one of the best works I have done in my career. I enjoyed the ownership, 0->1 projects & seeing the company scale.
• I was involved in the growth of the startup to 20x orders (~0.2 million per day), 5x Employees, ~$50M to ~$1B valuation (2017).
• Developing micro-service architecture, optimising the existing system and scaling the system.
• Real-time Fleet Tracking & Area Control: Built a real-time GPS ingestion and geo-zone aggregation pipeline to track delivery fleet health. Developed dashboards for SLA, backlog, and zone load. Enabled dynamic throttling (e.g. zone closures), reducing downtime by 25% and improving delivery throughput.
• Messaging Infrastructure: Re-architected Swiggy’s central message broker, RabbitMQ, from a single node to a highly available cluster behind HAProxy—eliminating a critical point of failure.
• Experimentation Platform: Built the internal A/B testing platform used org-wide to run experiments. Included a logging framework for writing experiment data to Amazon Redshift for analysis.
• Scalable APIs & Latency Reduction: Developed v3 of the core food-ordering API, handling 200K+ orders/day. Designed an application-level caching strategy with Redis, halving API response times.
• Fallback Distance Service: Built a distance service using OpenStreetMap as a fallback for the Google Maps API, reducing dependency costs and improving system resilience.

Aug 2014 – Sep 2015 · 1 yr 1 mo · Bengaluru Area, India

• Part of engineering enablement team in cloud and virtualization group.
• Developed products which helped teams to manage their issues well.
• using Groovy on grails, JQuery, Hibernate, Mysql.
• Implemented continuous integration pipeline using Jenkins

Dec 2012 – Dec 2012 · 0 mo · Mumbai Area, India

• Headed the National level chess competition as a part of sports festival of the college.
• Manage marketing participation, judges and related concerns of the event with a participation of more than 100 students across Maharashtra.
• Increased the participation in event by more than 200%.