Hello There 👋, I am a
25 y/o Software Engineer


Currently:

Building solutions with GenAI @ CIBC
Making AI videos at ParagonAI
-

Building and deploying GenAI solutions at CIBC, focusing on LLM-powered document processing, RAG systems, and agentic AI for internal banking teams. Collaborating efficiently with LOB stakeholders to define requirements, perform rigorous evaluations, and ensure high-accuracy outcomes. Work with AI scientists to develop secure APIs using FastAPI and LLM workflows using OpenAI and open-source models.

-

Collaborated with data scientists to develop and implement generative AI projects, focusing on enterprise business use cases and data insights with the latest LLM models. Supported front-end team in SaaS product development using ReactJS and created Python scripts for processing large conversational datasets. Built a custom RAG chatbot deployed on a GCP VM. Handled relationships with an enteprise client portfolio and managed key objectives to ensure successful project delivery and SaaS usage.

-

Built a responsive React chatbot application with a complex RAG architecture for extensive unstructured PDFs. Built a Flask backend on a GCP VM with secure JWT authentication and efficient logging for an Admin Portal, enhancing customer usage metrics visibility. Achieved recognition from Llama-Index and garnered over 10,000 views on YouTube for work with Pinecone and OpenAI embedding models.

An agentic application to help autonomously analyze and provide useful suggestions regarding personal habits using tool connections with Notion and Google Calender. Posting as a series on YouTube.

Building a custom SaaS application for AI-driven roleplay chats using open-source and fine-tuned LLM models hosted on third-party cloud platforms. Creating a responsive, performant front-end experience with TypeScript, ReactJS, TailwindCSS, ShadCN, and ReactQuery to deliver a seamless chat experience. Implementing a robust backend with FastAPI to build asynchronous RESTful APIs for seamless integration with Supabase and various LLM providers.

Build and deployed a production AI chatbot RAG application with a complex RAG pipeline of 40 unstructed PDFs. Custom made front-end with ReactJS and Tailwind, with a back-end Flask server to manage authentication, chat context and AI functionality.

For final year Capstone Project, developed a machine learning based Personal AI Gym Assistant application. Developed non-linear ML models using TensorFlow MoveNet to analyze exercise video data. Used random forest regressor model and gradient boosting regressor from SciKit Learn package based on angle data from MoveNet to predict form. Developed valuable user reccomendations based on live analysis of body key points to reduce injury and maximize growth. Watch the breakdown on YouTube!

A web-based complete interactive travel advisor application that is built on React and uses multiple GoogleMapsAPI and TravelAdvisor RapidAPI to recommend great restaurants, attractions, and hotels to users. The website is fully dynamic/responsive and allows users to sort different places based on their ratings. It also displays useful information such as any awards recieved, star ranking, address, and more.

A web-based space themed social media app built on React using NASA's APOD API. The images are automatically updated based on the current date and older images are moved down chronologically. Functionality includes ability to filter images based on specific date range, like and unlike as well as request further information.

A full-stack branch management application for a video game store. This project was built using a React front-end, Express backend, and a MySQL database. Functionalities include the ability to insert new employees, ads, or stock. The view displays multiple complex queries. Recommendation uses previous data to sell new games. Finally, accounting functionality displays useful branch revenue details.

Want to get in touch? Contact me on any of the platforms!