Varun Jewalikar

I am a senior software engineer with 15 years of experience - 9+ of them at Amazon (Prime Video) - leading teams / orgs, setting the multi-year vision, shipping production distributed systems scaled to millions of RPS and working across 100+ internal teams as stakeholders. I recently shipped an AI code-migration product that saved years of dev time for 50+ Amazon teams.

London, United Kingdom

Senior Software Development EngineerAmazon
  • Bootstrapped the product to automate code migrations for Prime Video (For example - CloudFormation to CDK migration). Designed and built a production LLM-powered code-transform tool - prompt orchestration & re-use, custom tools, context management. Cut time per migration from weeks to a day, saving years of developer time across 50+ teams.
  • Led the multi-year architecture, strategy and built the team owning Prime Video's edge routing layer - millions of RPS, >four 9s availability, <1ms average latency, 8 AWS regions, 100+ internal teams as customers. Significant improvements to Prime Video CX and architecture - stateless routing, rule based routing, zero touch automated route changes, scheduled route changes, safe rollouts of routes, regional failover and more.
  • Designed and implemented AWSSSMChaosRunner, a chaos testing/fault-injection framework used by 200+ teams internally. I shipped/maintained this as a solo dev with contributions from SDEs across the company. Presented at AWS re:Invent 2020.
  • Shipped a DynamoDB global replication system outperforming the native offering - improved uptime from 90% to 99.999% and replication latency.
  • Rearchitected the cache eviction layer for Prime Video's location service, cache hit rate went from 80% to 99% saving millions/year in cloud costs.
  • Presented popular tech talks twice at DevCon (the annual internal developer conference) and once at TestUnconference.
  • Mentored 15+ Amazon engineers across levels. Active interviewer for London SDE loops. Mentor for student outreach programs.
  • Authored white papers for the AWS Blog about chaos testing, Kotlin adoption, etc.
  • Authored Amazon's official manual for writing technical design documents.
MentorCodementor
  • Highest rated mentor on the platform for Python/NLP/ML, with hundreds of mentees - professionals and students.
Founding EngineerStealth Startup
  • Designed/implemented the image vectorisation system to detect affiliate products in user images.
  • Rearchitected the database layer (90% latency reduction).
  • Second engineering hire on a four-person founding team, building an image-based social network for fashion shopping.
Founding Research Engineer and Developer EvangelistMusixmatch
  • First R&D hire. Bootstrapped the research platform for the world's largest lyrics database — timestamp alignment, lyrics-based music recommendation, automated improvements of user-annotated lyrics, etc.
  • Published viral data studies featured by news outlets, such as The Guardian. Presented the company at 20+ hackathons across the US and EU.
R&D Engineer TraineeYamaha Corporation
  • Shipped algorithm from my master's thesis into Yamaha's Vocaloid product. Achieved a 20% improvement in time alignment of data.
Research InternMusixmatch
  • Prototyped NLP and ML systems to improve lyrics data quality and CX.
Google Summer of Code InternMixxx DJ
  • Added key detection, tempo locking and harmonic mixing features to Mixxx, an open-source DJ application written in C++.
Google Summer of Code InternBeagleBoard.org Foundation
  • Developed a userland/kernel-level driver for the PWM capabilities of the BeagleBoard single-board computer by Texas Instruments.
Master's Sound and Music Computing
Universitat Pompeu Fabra · 2011 - 2013 · Barcelona, Spain
Bachelor's Computer Engineering
Delhi College of Engineering · 2006 - 2010 · Delhi, India
Java / KotlinAWSTypeScript / JSPythonAI-native productsRoadmapping & org chartersPostmortems & incident managementPrototyping & productionisingCareer coachingMentoring at scale
Download