Building the next Financial AI Product

- Led Revenue Optimizations Tech, implemented multiple strategies for ad-trading i.e cross-bidding, pre-bidding.- Delivered 10+ features for clients and revenue tracking dashboard, major feature being calculating real time earnings, settlements with Ad Exchanges.- Revamped and Implemented whole AWS Infrastructure, re-designed and reviewed the HLDs and LLDs for better implementation in terms of CI-CD cost(build times), exhaustive testings(unit-testing and integration-testing) and developer hours to implement and push a feature to production.- Brought the overhead expense of development + product team by 50% and cut short the goal of engineering profitability by 3 months.

ByteLearn is the AI math teaching assistant for teachers and students, enabling teachers to support every student with 1-on-1, step-by-step help. With ByteLearn, differentiating practice for a whole class is instant and easy, and the system automatically finds and fills students’ knowledge gaps.- Build ML/DL models for generating step-by-step solutions for mathematical questions.- Lead Development of ML Powered User Facing Product.- Replaced 3rd party framework with an in-house framework with better results in terms of latency, RPS, and explainability.- Build end-to-end user-feedback/bug-catching pipeline to automate the process from ticket filing to RCA, removing manual dependencies, and saving countless hours in log processing. - Improved in-house data collection pipeline, streamlined flow between SMEs and Software Engineers for flow of data, implemented multiple checks to ensure quality, consistency and reliability of data.

- Helped multiple clients discover domain-specific AI-based use cases, and engineered solutions to work in real-life scenarios.- Build multiple DL models on top of CV and NLP to serve use cases such as Industry Level Fire Grading, Panic, and Feeble Sound Analyzer.- Build optimized models to deploy on edge devices such as NVIDIA Jetson Nano and Raspberry PI.

Hello-World is a non-profit organization based in New Delhi, India. The idea of hello world is to provide free, industry-aligned talks and insights. Hello-World is heavily invested in the upliftment and education of people from economically troubled backgrounds.- Taught Classes for Development, BootCamps, and Machine Learning.- Part of organizing team the first-ever 24hr+ hackathon of the College for free of cost with world-class Mentors and Guides.- Part of the team that brought Hello World on radar in New Delhi. - Organized 10+ workshops, and seminars on multiple emerging techs such as Machine Learning, Cyber Security, AR, and Unity Game Development.

IIPC acts as a gap closer between the industry professionals and the students of GTBIT. It aims at providing world class exposure to the persons with everlasting desire for doing something impactful.

Contributed to Groot Personal AI under the GSSoC'19 umbrella. Project tech stack: machine learning, web scraping, python, google apis.Contribution : 7 pull requests merged (OpenCV, Machine Learning, Pyqt5)