I am a graduate from IIIT-Hyderabad where I studied Computer Science. I have an interest in high performance computing and large scale distributed systems.
I have worked on the following components inside https://www.palantir.com/platforms/foundry/
1. Worked on the re-architechture of the metrics backend for foundry, made query p99 to go from 1.5sec to 300ms. Added support for aggregations and aggregated monitoring in the platform.
2. Implemented a concurrency control mechanism in foundry's job orchestration backend to auto-detect concurrency limits for downstream services in order to achieve optimal throughput with optimal latency
3. Contributed to a re-architecture of Foundry's internal event queue mechanism ; interfacing with the storage stack down to the transactional layer https://github.com/palantir/atlasdb; 10x less load.
4. Added support for checkpointing in the event sourcing mechanism for scheduling service. Allowed us to remove long standing bugs which saved customers money in accidental builds. Reduced service startup time from 1000s to 10s. Enabled development of newer features to be built on top of checkpointing.
5. Re-writing the internal event broker service to allow for more scale.
6. Led database migration for job orchestration service , architecting schema redesign and caching
optimizations that reduced database read rates by 85%. Designed cross-database transaction coordination and event
bridge layer enabling zero-downtime cutover.
If you find these interesting, come work with us https://www.palantir.com/careers/