Projects

OpsAgent

Developed as a functional MVP for OpsReady, an industrial management firm in BC, OpsAgent is an AI-driven assistant that simplifies navigating complex work order systems. Moving away from traditional search, I collaborated in a three-person team to engineer a Tool-Calling Agent using Claude’s API. My contributions focused on building the backend logic for 14+ custom tools, allowing the LLM to execute real-time queries and perform actions across the company’s operational infrastructure through a natural language interface. Python, FastAPI, Claude API, React, REST APIs

Tech Stack: Python (FastAPI), Claude API (Anthropic), Tool Calling (Function Calling), and REST APIs.

LinksV8

Currently in active development for the Kamloops Art Party, LinksV8 is a community-matching platform engineered to bridge the gap between digital discovery and real-world social connection. The backend features a robust PostgreSQL relational schema optimized for geo-spatial discovery through PostGIS, utilizing geography data types to power proximity-based event filtering. While the current sprint has finalized the core UI scaffolding and secure Supabase Authentication, the upcoming roadmap focuses on implementing a sophisticated Interest-Vector matching engine and real-time WebSocket messaging to automate the formation of pre-event group chats.

Tech Stack: Flutter, Dart, PostgreSQL, Supabase, PostGIS

Cinemax

CineMax is a personal movie discovery platform inspired by a night spent with a friend struggling to find high-quality retro films in a sea of modern blockbusters. Designed to bridge the gap between niche tastes and mainstream catalogs, the app uses a Hybrid Recommendation Engine that combines Collaborative Filtering (SVD) with a custom Content-Based boosting logic to surface hidden gems. Currently serving a close-knit circle of film enthusiasts, CineMax offers a sleek, dark-themed experience for building curated playlists and exploring cinema by specific eras and genres

Tech Stack: Python (Flask), React.js, NumPy, Pandas, SciPy (SVD), PostgreSQL, and JWT Authentication.

FeelScape

FeelsCape is an AI-powered music generator designed to create distraction-free LoFi soundtracks for studying and deep work. The system uses computer vision to analyze the color temperature and composition of user-uploaded images, synthesizing custom ambient tracks that match the visual mood of their environment. While currently focused on generating short-form loops for a small user base, the project is expanding into a generative platform capable of producing endless, personalized audio streams for productivity.

Tech Stack: Python (FastAPI), OpenCV, NumPy, Scikit-learn, MIDI2Audio, and React.js.



Contributions

SecureAI - Open Source Contribution

Contributed technical documentation to SecureAI's vulnerability testing framework. Researched and documented 7 vulnerability categories, 20+ probe types, and corresponding detection mechanisms for LLM security testing.