Jan 2021 – Nov 2022 · 1 yr 10 mos · India

• Worked on productionizing various ML Models to put into production (recommendations, classification etc.)
• Scaled various existing services to be able to handle upwards of 50 RPS by using ASGI and WSGI serving capabilites.
• Designed and implemented Spark based ETL pipeline for doing massive computations for creation of feature sets (used by DS for creating models)
• Designed and implemented workflow management for the same, to keep track of various pipelines.
• Designed and implemented various ML models for Streaming data (Spark Streaming)

Dec 2019 – Jan 2021 · 1 yr 1 mo

• Designed and developed, end to end, a schedulable, scalable, configurable ETL Pipeline system that makes it possible to use real-time events to construct clean, aggregated data using the Airflow and Spark DAG systems (plus Lambda, Kafka, S3, Redshift)
• Enabled easier independent scale-up of the Product and Analytics suites through eventual decoupling of Analytics and Tech databases
• Created the main ETL Scheduler, various Lambdas (to receive and parse events from clients and webhooks), the event emitter frontend library and the Flask service that interacts with our data warehouse (for custom admin/analytics dashboards + Alembic)

Jun 2019 – Dec 2019 · 6 mos

• Designed and developed an Analytics and Reporting System for teachers to view students data, from scratch using MongoDB, Flask and ReactJS.
• Revamped the process of curriculum graph design by Designing and Developing a curriculum-graph frontend system to enhance the ease of use and speed, using a drag and drop interface
• Mentored a team of 3 to create math animation videos, analytics dashboards (using python scripts and google sheets API) and an internal reporting tool powered by analytics
• SPOC for product and curriculum specific data requests.

May 2017 – Nov 2018 · 1 yr 6 mos · Greater Chennai Area

• Worked with Prof. Sitabhra Sinha from the Complex Systems and Non Linear Dynamics team.
• Analysed a globally coupled Ising Spin system with Hebbian Learning in a Multiplex Network; Direct application in Pattern Retrieval from Hopfield Networks (Undirected Neural Networks)
• Presented a model on Iterated Prisoner’s Dilemma on a network with opponent awareness (Memory of opponents’ strategy)
• Working on the Spotify Million Playlist Dataset using sparse graph representations for creating a playlist recommendation engine.

Aug 2016 – Jan 2017 · 5 mos

• Designed strategy and execution workflow for 'Jan Shikayat Ki Panchayat' campaign targeting 7.5 million people nationwide
• Headed Kisaan Door to Door Campaign for INC for upcoming Uttar Pradesh Elections in 3 districts: Shrawasti, Ghaziabad, Amethi; Total outreach of over 1000000 farmers
• Managed Logistics, Execution for Rahul Gandhi Khat Sabha in Ghazipur, Ambedkar Nagar

Nov 2015 – Jul 2016 · 8 mos · Greater Delhi Area

• Redesigned Spring, Hibernate based Backend to integrate Direct Messaging feature with GCM, in app-payments
• Mentored 2 interns to create a model to extract user’s dating preferences from facebook likes, shared posts, friends’ preferences

Sep 2015 – Jul 2016 · 10 mos · Gurugram, Haryana, India

• Worked on creation of a gamified social network that connects like-minded travelers
• Responsible for product architecture, design, implementation and placement in market
• Tested product concept and collected user preferences for product by accessing travel
• communities as Head of Operations for two socio-travel startups: india42, wanderwagon

Apr 2015 – Sep 2015 · 5 mos · Gurugram, Haryana, India

• Designed price discovery models for derivative pricing in the Indian Stock Market
• Implemented time series (ARIMA, GARCH) and portfolio theory based models (CAPM, MPT) and backtested various algorithms to check periodicity, basket arbitrage possibilities
• Used various technical indicators and basic greek analysis to trade options and futures.
• Built analytical tools using R to extract relevant information out of large datasets

Dec 2014 – Mar 2015 · 3 mos · Gurugram, Haryana, India

• Designed management system that predicts person-task fit based on skills and past performance
• Implemented supervised learning techniques (Regression, NLP) to predict time taken to project
• completion using past performance data, skill tags, human learning curve

Jul 2014 – Dec 2014 · 5 mos · Gurgaon

• Currently work in the client services wing of SGPM group at IMS health. Have executed Social Media Analytics based Service requests as well as those related to MP reports.

Feb 2014 – Feb 2015 · 1 yr · Greater Delhi Area

• Was responsible for event management on ground as well as off ground, heading core operations as well as marketing at india42 as a part time job.
• india42 organizes experience based trips to offbeat locations across the country.

Aug 2012 – May 2013 · 9 mos