* Identify SQL server Database blocking to handle the scale of 1 Million jobs per week running on an in-house built orchestration engine and reduce write and read latency by ~60% to and from critical production tables. ***( Please refer to the Project section below for more details)**** Subject Matter Expert for 1. Dataproc clusters to run machine learning models using spark and hive 2. Watchdog for moderating the job queue and its availability and scale, and 3. Health check service to monitor multiple microservices in the distributed system for higher availability.

• Managing the full lifecycle as a DevOps engineer from development, testing, deployment, scalability, and monitoring of our Wayfair-built Distributed job scheduling system called Kronos. • The system is responsible to run python, SQL, bigquery, dataflow and dataproc jobs on a schedule and ad-hoc basis. • The system is fully migrated and deployed in the Google Cloud Platform and runs approximately 1 million jobs per week in total with over 1000 unique users and around 250,000 unique procedures.

• During events like Cyber5, created automated tool using Python ,datadog API to quickly decide how much to scale up 25 Kafka clusters with 3000+ nodes retaining ~ 600 TB of data across 5 Google Data centers with cost and resources in mind to keep up with the average increase in site traffic by ~ 60% ***( Please refer to the Project section below for more details)*** • Automate terraform code changes to scale 25 Kafka clusters across 5 Google Data Centers up and down to save time from 2 weeks with 4-5 engineers to less than a week with 1 engineer.• Automate kafka topic expansion to scale high volume topics across 3 major stakeholders to prevent lags in consumption and data loss. we support around ~24M+/sec messages in Kafka across all topics.• Supporting the hybrid cloud Kafka system to guide the stakeholders for efficient consumption of streaming data.• Maintaining and updating the system using Terraform.• Batch load Jira API data to Big query table using Apache Beam with Google Dataflow to help project managers to optimize sprint tasks.

• Consult with BI and engineering teams to build data products and expand our understanding of the performance and usage of our systems. Designed neo4j graph database of around 4000 nodes and 13000 relationships to track dependencies between different entity types in the system for root cause analysis.• Leverage existing and create new data pipelines from a variety of internal systems and build data models off of those data streams by cleaning and transforming using regular expressions (regex).• Interact with Google Cloud Platform (GCP) using their python APIs to stream Big query logs in JSON format using the Stackdriver API. Configure and setup production GCP resources and assign custom IAM roles using terraform.• Deploy ETL on the Google dataflow for its auto scaling and distributed processing features using Apache Beam to reduce the run time significantly.• Coordinate development and deployment among several engineers across multiple applications using GitHub.• Provide training and mentoring to team members in areas of expertise as needed• Design and develop maintainable, performant Python and SQL code to drive our internal tracking datamart and processing of 11 platforms with around 700 million records and growing by 5% each day. Applied session logic of 2 hrs for user profiling.• Participate in code review discussions within the area of expertise• Build an understanding of our applications across the entire stack, from server configuration to our backend data processing application to how that information is exposed and analyzed• Perform database design, development, performance optimization, ETL in a variety of SQL environments, including both traditional RDBMS and MPP systems such as HP Vertica and Microsoft SQL Server

• Built a neo4j NoSQL graph DB applying ETL for historical corporate hierarchy data from FactSet terminal for 3000 companies for index creation using existing patent data ***( Please refer to the Project section below for more details)***• Transformed the Postgres relational (RDBMS) to Neo4j graph NoSQL database applying ETL due to citable nature to improve the response time to at least log(n). In the transition to deploy subgraphs into hadoop infrastructure with multiple data nodes for parallelism, replication and saving cost by horizontal scaling on AWS EC2.• Designed, architected and implemented to harmonize different variants, pivoted aliases to family corporate entities from 58 million dataset using Python and linux bash. To track innovation of all subsidiaries within a family e.g Apple.• Developed using statistical methods to rank companies on 4 weighted metrics in R, created tableau dashboard (see tableau public profile) to control parameters dynamically and applied K-means clustering to have the replacement for the companies.• Follow Agile to set up the proprietary patent system with 250 million data points on AWS (EC2, S3, RDS, Docker) cloud using Github for dynamic load balancing and 99% uptime. Linux command line used for deployment.• Saved approximately $6000 over a year on the AWS account by using CloudWatch and cost explorer.• Developed a secured web portal in LAMP Stack storing structured data with Integral Accounting framework of 6 dimensions for decision making using PowerBI.• Built a VBA Macro tool to reduce the time by 1/3 for stock selection process for trading and perform back testing for historical performance.

• Assisted the Professor with his undergraduate Web Application Development course which covers HTML5, CSS, PHP, ASP.Net• Graded 35-40 students per semester and provide constructive feedback to improve their performance.•  Modeled Use case and Entity Relationship Diagrams for 14 academic web projects for high level functionality and database schema.

• Serve walk in or remote assistance to CIS undergraduate students for their assignments,course work, projects or conceptual queries in VB.NET, Data Mining, Linux, System Analysis & Design.• Manage help desk schedule and reporting trends related to number of incoming students and their queries which are associated to Professor Dan Turk.

• Performed Business Intelligence using Power BI to analyze status of test cases, time in scrum meets to improve automation performance• Followed Agile process to automate test cases written under gherkin in C# integrated with Selenium,Ranorex to save time on manual testing • Integrated Selenium Drivers with Python to automate the web-based automation features for open source flexibility and faster response time• Automated 56 scenarios with 237 test cases across 9 features integrated with Nunit installed in 3 VMs to reduce turnaround time by 1/3• Merged different versions of code of the team using Git Hub Repository and tracked with TeamCity for reliable maintenance of the software

ALGORITHMIC TRADER : Developed systems for the Proprietary desk are as follows : • Exchange approved High Frequency Trading (HFT) systems which traded about USD 20,000 worth of value in cash and futures on weekly basis and saves 90% of trader’s time per day. • The Index Arbitrage for multiple indices such as NIFTY, BANKNIFTY which trades about USD 1.6 million worth of value on weekly basis reducing pricing errors due to the high speed execution on order entries. • Market Making Algorithms across multiple exchanges like NSE,BSE, MCX for Exchange Traded Funds (ETFs) INSTITUTIONAL QUANTITATIVE ANALYST :Developed and implemented automated models:• Updated in real time 10 emails of corporate announcements of the derivative stocks, to be the first to inform clients.• Predicted daily 3 series of events in dividend announcements with the impact on the indices to take a call on open and fresh trades.• Predicted price movement event wise for stocks under merger and acquisition to enter a pair trade.• Updated in real time the traded client wise stocks in a client friendly format to speed up their decisions.• Monitored the sector wise trading calls from 20 analysts with daily notifications to the analysts and their respective reporting managers for validation.• Summarized view of the Indian stock market during the expiry week for volatility and market wide positions across brokers and clients.• Updated daily the 50 ETF flows of US funds in Indian equities with daily and weekly changes stock and sector wise for the performance of the Indian stocks in the foreign sector.DATABASE ADMINISTRATOR :• Maintained and Updated the MIS System to generate daily and monthly reports as per requirements using SQL to foster intelligent business decisions• Developed a Risk Management System which monitors various parameters to ensure Risk free Trading using VB.net and SQL which saved huge losses during volatile markets.

• Automated diurnal activity (Billing details,Patient details and Sharing of Funds to Doctors and Trust) across 5 medical Departments to evaluate more accurate finance and mitigate manual intervention. • Considered the Information Security Factors such as Confidentiality,Integrity and Availability of Data for the implementation of the System.• Provided Business Continuity with replicas and auto daily backup of the Database for disaster recovery.

Developed a Digital Magazine App on the Android Platform to be downloaded from the Google Play.

• Developed Games for Web Browser and Platforms such as iOS and Android ,HTML5 , JavaScript and C ++ with the Licensed Marmalade SDK