Sep 2016 – Jun 2019 · 2 yrs 9 mos · Hyderabad Area, India

• ServiceNow has lots of infrastructure automations running and it needed to know estimated time of its completion. Did a POC and suggested management to use machine learning for solving this problem. This project required in depth understanding of different infrastructure automations, ServiceNow platform, Scalable Design, Machine Learning. Created whole pipeline for machine learning model creation and usage. Made it scalable, fault tolerant, easy to monitor for different models.
• Made it as a framework so that it becomes very easy to onboard any automation to start getting estimated run time.

Mar 2014 – Aug 2016 · 2 yrs 5 mos · Hyderabad Area, India

• Added ASR (Azure Site Recovery) Initial Replication progress feature, which is showing Initial Replication progress in the ASR portal for VMs. It basically is getting replication percentage by calling WMI calls to VM in scalable manner and sending it to ASR service which persists in its database and Portal uses that to show it to user.
• Worked on Process Server failover feature, which is to failover Process servers. Process server are onprem gateways for the replication of the VMs to Azure. As this onprem and it resides on a VM it was a single point of failure which results in breaking of all the VMs. With PS failover feature a new instance of PS can easily be created in case of its failure. It involved developing UI for taking failover inputs and transferring the load at backend.
• Worked on many features in Azure Site Recovery related to failover, replication, licensing etc.

Aug 2012 – Feb 2014 · 1 yr 6 mos · Hyderabad Area, India

• Worked on development of Analytics Reporting website for inventory management internal to Amazon. I was involved in its design and development of both backend and frontend. Developed scalable website using AWS technologies (such as EC2, S3, RDS, Dynamo DB) and GWT.
• Created a message processing and storage system to handle millions of messages received in SQS (Simple Queue System) daily. The messages get parsed and useful information is extracted and stored in Cassandra and then moved to hive for performing queries on it. Wrote a parsing logic to parse the JSON data (by converting it first to Map of Maps) and extracting useful fields by using a configuration file. Then the useful information gets stored in Cassandra in real time. Cassandra column Family data is extracted to file, which then gets loaded to hive for queries.
• Worked on an alarm detection system. It used different thresholds provided in configuration file and used it to detect any alarm. Once alarm got detected it sends mail to configured mailing lists. The mail is well formatted in showing the Alarm Data.

Sep 2010 – Jul 2012 · 1 yr 10 mos · Bangalore, India

• • Developed a Library, which provides API’s for secure login of Tally users to TNS (Tally Net Server). This involved deep understanding of socket programming, securing the library and making it difficult to de-compile, secure socket layer, Model View Controller, ways to decompile.