M.Eng. Student · Software Engineer

Making AI faster, smarter,
and more accessible.

Torrin Conrath

AI Confidence Scoring Framework

A comprehensive framework that assesses and quantifies the trustworthiness of AI-generated outputs from RAG-based enterprise knowledge systems. It employs a multi-faceted confidence engine, combining consistency, sensitivity, and novelty scores, to provide a unified, explainable reliability score for AI system outputs, enhancing decision-making and transparency.

I worked in a team, specifically, I developed the backend RAG and logging systems. This includes building a FastAPI service that ingests documents, vectorizes them, and then uses them in the confidence engine to score the user query's response. The query results are served to the React frontend, and all activities are logged throughout the process in a PostgreSQL database.

Mugen AI

An AI-powered personal productivity app generator built at HooHacks 2026. Describe any app you want — a habit tracker, spending dashboard, mood journal — and Mugen builds it instantly as a fully functional React app, pre-filled with your own real data. Paste in raw text and Claude infers your schema, links your data sources, and ships a complete app in seconds.

Jerry The Medical Assistant

A fine-tuned medical chatbot built on Qwen-8B, designed to simplify complex medical jargon for older patients. Fine-tuned using SFT and LoRA on 20,000 doctor-patient dialogues, accelerated with vLLM and AWQ quantization, and evaluated with a 48.5 SARI score and a 4.5/5 usability rating.

Robust Pest Detector

Real-time raccoon and rodent detection using lightweight YOLO models, deployed on edge hardware. Trained and compared YOLOv8n, YOLOv11n, and YOLOv11m across two custom datasets, stress-tested with synthetic augmentations, and deployed to a Raspberry Pi and desktop via a React dashboard.

Car Score Predictor

A full-stack mobile app that scores used car listings to surface real deals. Type in a car's details in plain text and get a data-driven value score back. This mechanism is powered by a neural network, an NLP parsing pipeline, and a live database fed by a web scraper.

Agent vs Ghosts

A top-down wave shooter built in Python and Pygame ported to HTML+JS. Dodge and shoot endless waves of escalating ghosts, collect 8 powerups, and survive as long as possible. Playable in the browser.

I'm an aspiring software engineer and M.Eng. student at Virginia Tech, developing solutions with an actual purpose. My recent work reflects my interest in making AI more lightweight and accessible across a larger number of consumers. I'm focusing on developing full-stack applications and systems that run close to hardware, from training models to shipping interfaces for specific user needs.

torrinconrath@gmail.com ↗

Stack

  • LanguagesPython, JavaScript, TypeScript, C/C++, SQL
  • FrameworksReact, React Native, Flask, FastAPI, Expo
  • ML / AIPyTorch, Hugging Face, YOLO, OpenCV, vLLM, LangChain, Quantization, Transformer Models, LLMs, TFLite, ONNX, Tesseract.js, Claude API, ElevenLabs
  • InfrastructurePostgreSQL, MySQL, ChromaDB, Docker, AWS, Vercel, Supabase, Ngrok, WSL, Git, Unix, CI/CD, Github Actions, MLOps
  • HardwareRaspberry Pi, Edge Devices, Servers