- Leading the development of a Cash Flow Forecasting System for one of the leading Investment Banking firms across the world. This helps the firm make key decisions on their investments based on the forecast of historical data and/or how it differs based on different scenarios like changes in tax obligations, sudden economic crashes, anticipated change in specific inflows or outflows and so on.- Developed an intelligent document processing system that takes in a bulk of documents and finds specific keywords or phrases whilst also identifying or classifying what each document could be and optionally giving an option of running extraction processes on pre-defined document types and also running key validations on the data extracted by the documents.- Managed end-to-end AI capabilities as a Client Engagement Manager, leading tech engagements and developing key applications, such as extracting information from ITRs, NACH mandate documents, and cheques for one of India’s leading Insurance companies.- Built a multi-lingual speech-to-text module that transcribes audio and matches important keywords and phrases from a pre-defined script, enabling clients to create an ops-free insurance purchasing system.- Conducted interactive webinars for global clients to demonstrate how to leverage our API Marketplace within their internal systems, leading to time, cost, and resource savings and improved business efficiency.

- Developed and deployed multiple data extraction APIs on the cloud for a major Indian bank, enabling data extraction from forms, IDs, and cheques from banks in India and the Middle East, as well as Dutch and German Visas. Led the development of computer vision-based APIs for extracting signatures, stamps, and photographs from documents, resulting in significant time and resource savings.- Designed and implemented a comprehensive system for information extraction from various Indian identification cards (Aadhaar, PAN, Passport, Voter ID) to streamline and expedite the KYC process. This solution enhanced efficiency and speed of KYC operations, benefiting 85% of all insurance companies in India by facilitating customer onboarding.- Built a suite of APIs for an API Marketplace, optimizing internal processes for banks and financial institutions, enabling faster customer engagement and increasing revenue.- Developed an API for audio call transcription and analysis, providing insights on sentiment (positive, negative, neutral), emotions (happy, angry, rude, neutral, polite), and mentioned entities (persons, organizations), and generating the complete dialogue in script form. This enhanced sales and customer support call tracking, improving employee coachability and response rates.

- Devised, Developed and Tested solutions for various clients across various industries in the New York region focussing on Computer Vision, Speech and NLP.- Suggested and devised solution architectures on Google Cloud Platform Vertex AI, applying complex customer requirements and best practices for deploying machine learning solutions.- Collaborated with customer technical leads to scope, manage, and deliver cloud-based Machine Learning, AI, and Data Science solutions. Provided dependable guidance to decision-makers throughout the engagement.- Constructed various technical decks for technical sales pitches to illustrate solutions to client's problem statements and help them comprehend the flow of the solution in a simplistic way.

Primary work consisted of automating manual tasks performed in an insurance company using Python and concepts of machine learning, deep learning, computer vision, and Natural Language processing. - Projects worked on:- Speech recognition techniques to build a module for Speaker Diarization and comparison functionalities using NLP to optimize Product Sales Calls made to potential and already existing customers.- Built a video KYC platform using techniques like facial recognition, Optical character recognition, Visual Analytics, etc.- Provided Machine Learning based services for one of India's leading Financial Institutions.- Build a pipeline to automate the process of extracting data from scanned documents using OCR, OpenCV and Python.