Mendel is a healthcare AI startup

Built an ETL to process AI pipeline and Data pipeline. Technologies involved: K8 Operators, Scala, Pekko streams, Pekko HTTP, Postgres, ElasticSearch, pekko-kafka connector (for ETL pipeline), Redis, Apache Spark.

Also was building and managing backend services for medical-abstraction and annotation.

Since, the team was very small I was involved in building, deploying and maintaining all the related microservices along with architecture of it.

This helped with securing funding for Series-A and later for Series-B.

The services evolved from pilot sized data to millions of records per customer.

Started back into Workday for a second time :).

But this time around, UI is into the mix for full stack.

Day to day work revolves around react js, rx js, scala, akka and Apache Spark.

Apixio is a healthcare startup and I had the opportunity to help build couple of applications from scratch in the short 6.5 months time-frame.

Built Workflow to orchestrate human flows. Also built an ETL pipeline to generate workitem for the workflow. Technologies involved: Scala, Akka streams, Akka HTTP, ElasticSearch (database for Workflow - used as caching store), Parser Combinator + json4s (to build Workflow DSL, JSON-T and Expression Evaluator), elastic4s (emits ES queries from custom DSL thru ParserCombinator), akka-kafka connector (for ETL pipeline), Redis (for locks and other state management).

Built Metric capturing system to capture business metrics. Technologies involved: Java 8 (core library), Cockroach DB (primary storage DB), antlr (query abstraction for the library), Redis (lock and state store)

Scala, Akka Actors, Apache Spark -- Microservice to digest click-streams to generate union–find for computing active-time.

Note: Joined in as Senior Software Developer and became Principal Engineer around early 2017.

Day today work involves building and managing infrastructure to generate online timeseries like reports using ~ 1.2 TB of logs generated per day.

Technology stack: Written in Scala using Akka (FSM / Graphs / ...) and Apache Spark.

Typical WorkDay: a. Write Dataframe and RDD APIs to generate parquet file out of flat files. b. Manage the generation of online report (hacked by Akka thru Apache Spark). c. Improving the performance of (a) and (b) depending on data growth.

Typical Afternoon: Research on upcoming version of Akka, Spark and other related technology stack. Learn to improve the efficiency of running clusters and so on.

General operations work involve maintaining a distributed ETL pipleline to enable ingestion of above mentioned data. My fingers had been itching to take the same mechanics to streaming data as they arrive.

Haven't put together time to get into the real analytics of ML yet!

Note: I started out as Software Engineer and became a Consultant Engineer (commonly referred to as Senior Principal Engineer) while I left EMC.

Worked for BPM and Case Management Engineering group.

Technologies used at work

• Java, Spring Framework, ANTLR

• Gradle, Grails, Groovy

Day to day work involves writing code to manage BPM engine & its features on the cloud. Along with implementation of Case Management Event and Access Control Engine for the same cloud platform.

Work history...

• Designed and implemented the core components of CaseAPI and the corresponding runtime engine for Eventing Framework and Access Control Framework.

• Built expression parser using ANTLR to parse, validate (type validation) and transform (to SpEL) . Used for Case Management (event system) and BPM engine (timers, transition)

• Involved in developing concurrency APIs for Process Integrator that acts as a poller & provides HA for various protocols (HTTP, FTP, Web Service, JMS, ...).

• Write demo grade POC Web application to showcase ongoing features using AngularJS or Grails framework.