Architecture for Scale/Data Architecture evolution and Security Policy recommendations in NSX Intelligence. Mentored/mentoring junior engineers and set unit testing standards/culture for reliable shipment of releases.

Founding engineer on NSX Intelligence Security Analytics and Micro-segmentation recommendations product.

Delivered first five versions of the product - micro-segmentation configuration drift and compliance checking.

Customer Discovery: Came up with innovative ideas like "group reuse threshold" for tunable group reuse during rule recommendations.

Requirements negotiation, work breakdown, prioritization, API design, service code, spark job pipeline code, some performance tracking and analysis and basic machine learning model lifecycle management.

Basic understanding of Deep Learning concepts , SGD, CNN - convolutions , RNN, Encoders/Decoders, Seq2Seq, Backbone networks/Attention and Transformers and basic PyTorch usage.

Decent understanding of LLM frameworks like langchain, llamaindex and the React Chain of Thought Agent model.

Technologies:-

Spark SQL,Structured Streaming, Lake house architecture, Delta Lake, MLFlow, Graph parallel processing systems like Graph Frames, Druid, Spring, Postgres, Spark Redis, Data Lake indexing.

I understand internals as well as uses cases and design of above systems.

Apple Delivery and Developer Productivity and Process Engineering.

Started from a vague requirement of "enable testing against iTunes in production" to working solution.

Then Created infrastructure to run integration tests against iTunes production (Apple Media Products) for the first time in its history. This required understanding internal Mescal Security protocol, building a REST API Mescal Protocol Server on top of a Hardware Security Module(hsm-service) to issue secure tokens to test writers. And then integrating with the a complex Test Framework/runner(jinglebird) written in scala/akka/play.

Developed other features in jinglebird (e-mail tests results on schedule) and new javascript test api's and managed the production infrastructure of the above using docker/kubernetes. Used Nashorn here.

Demonstrated POC for Cloud migration of iTunes Search Stack services using Containers/Micro services. Python, Bash, Docker, CloudFormation API; This resulted in executives green lighting cloud migration.

Developed a Canary Analysis prototype that made calls to internal telemetry platform(Hubble) to compare existing version and new version results; Used Spark/Akka/Scala.

Developed Spinnaker clouddriver plugin for internal deployment service and evangelized Spinnaker. Platform and Infrastructure team(PIE) adopted my work later.

Evangelization and best practice propagation of internal Platform services (PIE) to iTunes teams - Gatling as a Service, RioCI/CD.

Developed a Quality Service client for PIE Quality Service and used that to report Jinglebird test runner test results into PIE Quality dashboards.

VMware Networking and Security Business Unit.

Principal Developer and lead on Version 1 NSX Next generation Distributed Firewall Management Plane API that provides Firewall as a Service in NSX.

NOSQL Data Model design

API concurrency and fault tolerance using activities( Pat Helland's paper),

Smart Rule positioning

Strong unit and load test coverage and established a TDD culture in our component.

Minor Contribution to Firewall Central Control Plane Design.

Investigated literature and algorithms for fast insertion of Firewall rules in amortized time.

Solved latency issues with existing product, re-designed tenant API's and tuned DB performance and Hibernate issues in current product.

Filed a couple of invention disclosures.

Drove requirements on other teams for Firewall needs in a distributed system.

Mentored new and junior engineers.

Started as architect, transitioned to engineering manager and Incubated the software development team from scratch in VCE SantaClara reporting to the Director .

Created a fast deployment/auto-scaling solution by bringing together Spring tcServer, Hyperic monitoring, VSphere API's , Vblock converged infrastructure management ( UIM) and VMware cloudfoundry paas. This product brought Amazon elastic beanstalk + Amazon Auto-scaling style functionality to Vblock hardware.

Setup engineering processes and wrote the first PRD. Identified needed engineers/skillsets, wrote job descriptions, recruited 4 principal/2 senior engineers, one UI engineer, set goals/responsibilities/direction for incoming team members and executed on that vision to create a version 1.0 solution.

Did some individual contribution too by developing a REST API based blueprint server serving XML/JSON using spring mvc, spring security,spring roo and postgres database.

Developed a basic version of Admin Console for VCE Vision Vblock Systems Management Software using extjs. Did a prototype using bootstrap, ThymeLeaf.

Our team published the following paper in the course of this effort - http://www.vce.com/pdf/solutions/vce-alp-reference-architecture.pdf

http://chucksblog.emc.com/chucks_blog/2011/08/the-thorny-challenge-of-elasticity.html

TeleNav is a Location based service(LBS) SaaS company best known for providing internet connected navigation on mobile phones and cars. I acquired a good understanding of what is going on in Mobile eco-system, near future and long -term.

1) Held Engineering Lead and Architecture responsibilities for our Highly available/highly trafficked mobile internet application server. I understand scalability issues well. I designed our 6.0 and 6.1 server with the help of frameworks such as Spring to effectively serve millions of mobile clients. Created abstractions. Worked or led chunking/streaming efforts. Split our server into swim lanes.

2) Redesigned and partitioned our NavMap (Map tile cache) by effective partitioning/sharding, tuning Berkeley DB and monitoring various Linux operational parameters and disk seek rates. Understood eventual consistency, Voldemort, Cassandra, Dynamo and other highly available key/value stores as part of this effort.

3) Conceived, architected, designed and developed a POI recommendation prototype based on where users Drove To . Desgined Cassandra schemas/column families. Used Cassandra for storing real-time DriveTo POI information, Understood Google News personalization in this process.

4) Conceived, researched and made a case for using Chukwa for improving our Log collection process. Delineated the use cases of Chukwa vs the inverted index storage in Cassandra.

5) Redesigned our device launch process with hot device deployment drops to achieve true multitenancy for different carriers.

6) Also acted as a release engineer and pushed code to production. Also debugged production performance issues as needed.