I'm a Full Stack Developer and Master of Applied Computer Science student at Dalhousie University. I build full-stack web apps (React + Flask/FastAPI), cloud-native and serverless systems on AWS, and applied ML features with PyTorch and LLM APIs. I also spend a fair bit of time in Terraform and Docker, since most of what I build has to actually deploy somewhere.

I care about code that other people can pick up and extend without needing me in the room. Good structure and documentation that treats the reader like they have somewhere else to be.

Languages

Frontend

Backend & APIs

Cloud & DevOps

AI / ML

Databases

Tools

Recruiters drown in resumes. HireLens is a full-stack platform that takes the first pass off their plate: upload a job description and a batch of resumes, get back a ranked shortlist, search the entire candidate history in plain English, and generate tailored interview questions. Built with Flask, FastAPI, MongoDB, React 19, sentence-transformers, and Gemini. Packaged with Docker and shipped through a five-stage GitLab CI/CD pipeline.

Serverless AWS application that automates medical insurance eligibility checks end to end. Users upload a bill and get a verdict (Approved, Partially Approved, Rejected, or Needs Review) within seconds, with a plain-English reason attached. Built on Cognito, API Gateway, Lambda, Step Functions, DynamoDB, Textract, SQS, and SNS. Fully provisioned through Terraform and deployed via GitHub Actions.

Applied ML research testing whether Grad-CAM explanations actually reflect what a CNN uses to make its predictions. Fine-tuned a ResNet-18 on chest X-rays and compared three CAM variants using a masking-based deletion protocol. Key finding: faithfulness depends entirely on whether the model was right or wrong. Written up as an IEEE-format research report.

Half of software engineering is naming things. The other half is renaming them.