Disrupting traditional market research with large scale AI powered solutions.Currently working on: 1. Billion Scale Semantic Search.2. Qualitative and Quantitative Evaluation of Clustering.3. Stance Detection4. Serverless Compute for large language models.5. Effective Prompting for very large language models.

Extended transformer models and large language models to extract information from log files.

Teaching Assistant for 2 courses:1. CSI 4142: Fundamentals in Data Science.2. CSI 5388: AI for Cybersecurity and Applications.Day-to-day activities include:1. Conducting Lab sessions2. Helping students to carry out research and projects.3. Preparing and grading assignments.4. Setting up the cloud infrastructure for projects and assignments.

Recipient of a research grant worth $25,000 by IBM, MITACS and the University of Ottawa.Worked with Professor Paula Branco, Prof Guy Vincent Jourdan and Prof Herna Viktor on the problem of Information Retrieval from Log Files.Developed a system for automatic information extraction from log files and heavily working on Information Retrieval techniques like Named-Entity Recognition.Developed novel text augmentation techniques specific to log files.Focused primarily on deep learning techniques including sequence-to-sequence, transformers and language models.Performed ensembling on deep learning techniques and tried to maximize the throughput of information extraction from logs.This work formed the base of my thesis titled: "A System for Automatic Information Extraction from Log Files".

Python / Airflow Developer @ mPulse MobileFebruary'20 - October'20- Responsible for creating data pipelines / DAGs using Apache Airflow.- Heavily refactored the existing code-base to remove code duplicacy.- Optimised existing workflows and made them ~10x more efficient.- Responsible for building new features + fixing existing bugs in the system.- Responsible for fixing the existing vulnerabilities and code smells in the code-base.Tech-stack - Python, Apache Airflow, Flask, Postgres, RedisPython / AWS Serverless Developer @ ZepplinJune'19 - January'20- Employed FaaS (function as a service) using AWS Lambda.- Responsible for writing REST APIs and business logic using various design patterns.- Responsible for carrying out database design.- Active part of the product's design and brain-storming sessions.- Took ownership of writing unit-test cases and clocked code-coverage of over 75%Tech-stack - Python, Flask, Postgres, AWS Lambda, API Gateway, Cognito

Libraries and Technologies used: Python, Sklearn, Pandas, Keras, Numpy, Flask, Angular
- Worked closely with Digital Transformation Team and did a POC on Network Intrusion Detection.
- Developed a framework that prepared Machine Learning Ensemble models dynamically.
- Used Python, Flask, and Angular to develop an operations management web app.