Experience

Omnistrate

• Software Development
• 1 - 100 Employee

• Co-Founder / CTO

  • Mar 2023 - Present

  San Francisco Bay Area



Confluent

• United States
• Software Development
• 700 & Above Employee

• Staff Software Engineer

  • 2020 - Mar 2023

  Mountain View, California, United States - Led the infrastructure design and implementation of a project that increased the scale of Confluent Cloud clusters by 4x at a total data throughput of 20 GB/s - Lowered request latencies by 10x for Apache Kafka in Confluent Cloud - Worked on implementing a Broker Lifecycle manager for transitioning the Apache Kafka controller from a Zookeeper-based consensus model to a Raft-based consensus model - Pioneered a new way to measure service load using request latency to better represent… Show more - Led the infrastructure design and implementation of a project that increased the scale of Confluent Cloud clusters by 4x at a total data throughput of 20 GB/s - Lowered request latencies by 10x for Apache Kafka in Confluent Cloud - Worked on implementing a Broker Lifecycle manager for transitioning the Apache Kafka controller from a Zookeeper-based consensus model to a Raft-based consensus model - Pioneered a new way to measure service load using request latency to better represent the true load on the system and provide actionable inputs Show less



Cisco

• United States
• Software Development
• 700 & Above Employee

• Senior Software Engineer

  • 2018 - 2020

  San Jose, CA Distributed Ledger Platform development - Designed/Implemented a NoSQL backing store for smart contracts - Designed/Implemented a role-based access control within a smart contract using a graph database - Designed/Implemented a backwards compatible NoSQL schema registry to parse objects across versions on the ledger - Designed/Implemented a data-plane graph node for real-time ED25519 signature verification in VPP (https://fd.io/technology/#vpp) - Designed/Implemented a custom… Show more Distributed Ledger Platform development - Designed/Implemented a NoSQL backing store for smart contracts - Designed/Implemented a role-based access control within a smart contract using a graph database - Designed/Implemented a backwards compatible NoSQL schema registry to parse objects across versions on the ledger - Designed/Implemented a data-plane graph node for real-time ED25519 signature verification in VPP (https://fd.io/technology/#vpp) - Designed/Implemented a custom Docker network plugin to program VPP (https://fd.io/technology/#vpp) - Designed/Implemented a trusted infrastructure module that uses a distributed ledger for container orchestration (think auditable Kubernetes). Supports trusted/distributed Docker container deployments IPSec control-plane development - Building a control-plane for StrongSwan (open source IPSec implementation), leveraging the VPP network data-plane for scalable, highly-available and hybrid (on-premise + cloud) deployments Show less



Amazon Web Services (AWS)

• United States
• IT Services and IT Consulting
• 700 & Above Employee

• Lead Software Engineer

  • 2017 - 2018

  Greater Seattle Area Was a founding engineer for Aurora Serverless which accounted for 50% of the total new cluster creations for Aurora after launch (in its first year) Integrated process profiling using Linux eBPF for detailed performance tracing of the Aurora Multi-Master database stack Lead the effort in diagnosing and improving performance of the Aurora Multi-Master database stack

• Aurora (RDS) Software Development Engineer

  • 2015 - 2017

  Greater Seattle Area Worked on the enhancing the performance of the Aurora 5.7 compatible database across the stack to support feature parity with the Aurora 5.6-compatible engine Tasked with the design and implementation of the Aurora Zero Downtime Patching Process to enhance database availability by preserving database sessions across process restarts during database maintenance windows Tasked with the design and implementation of the Aurora Smart Read Selector for improving read… Show more Worked on the enhancing the performance of the Aurora 5.7 compatible database across the stack to support feature parity with the Aurora 5.6-compatible engine Tasked with the design and implementation of the Aurora Zero Downtime Patching Process to enhance database availability by preserving database sessions across process restarts during database maintenance windows Tasked with the design and implementation of the Aurora Smart Read Selector for improving read performance by up to 27% across the storage nodes and reducing read latency jitter by 75%



AmpArmor

• Entrepreneur

  • 2013 - 2014

  Bengaluru, Karnataka, India As a lead developer on XDA developers, engineered a CPU temperature control algorithm and developed a gesture-based wakeup on an Android device - Implemented a dynamic CPU temperature monitoring and control system for improved response in critical conditions as well as lower resource use during non-critical conditions and completed its integration into the Linux Kernel source of a Sony Xperia Honami, Sony Xperia Yuga and Google Flo (Nexus 7) - Implemented a CPU weighted hotplugging… Show more As a lead developer on XDA developers, engineered a CPU temperature control algorithm and developed a gesture-based wakeup on an Android device - Implemented a dynamic CPU temperature monitoring and control system for improved response in critical conditions as well as lower resource use during non-critical conditions and completed its integration into the Linux Kernel source of a Sony Xperia Honami, Sony Xperia Yuga and Google Flo (Nexus 7) - Implemented a CPU weighted hotplugging algorithm that manages CPU core states based on the usage pattern - Achieved up to 20 % lesser thermal throttling and up to 40 % longer standby times on comparable platforms - Achieved up to 20% more responsiveness Show less



Bharat Heavy Electricals Limited

• India
• Appliances, Electrical, and Electronics Manufacturing
• 700 & Above Employee

• Intern

  • Jun 2013 - Aug 2013

  Bengaluru Area, India Algorithm and UI design for the Human Machine Interface (HMI) that simulated an Automated Pump Selection system for use in a Steel Plant at the Automation and Drives Laboratory, fully compliant with PLC programming paradigms