Hi, I'm Matthew McDowall
I build end-to-end AI systems that ship to production — not just demos. From RAG pipelines and LLM tooling to data infrastructure, I turn ideas into real products. Got an idea?
About
I'm Matthew McDowall — a Senior at the University of New Hampshire who builds custom, domain-specific, end-to-end AI applications. Not demos. Full systems that go from working with clients and concepts to production.
Currently interning at North Light AI where I architected a supplier-matching system adopted by 50+ SMBs and NASA. I'm deep into applied AI and LLMs — and making them useful in the real world.
Exploring models, datasets & spaces
Tech I Work With
Things I've Built
Prime Ready — NHADC
Hybrid RAG-based matchmaking system that pairs small-to-medium suppliers with Prime contractors, replacing 5-6 weeks of manual discovery with near-instant query responses. Adopted initially by 50+ SMBs and NASA.
AI-Driven NMS Chatbot — SubCom
Locally-integrated chatbot for SubCom's subsea cable Network Management System. Parses proprietary NMS documentation, embeds it into a vector database, and exposes natural-language querying through a full backend + frontend stack — fully air-gapped with zero external API calls.
Coastal Research Pipeline — CCSCR
Automated data-processing pipeline built for the Cohasset Center for Student Coastal Research. Handles weekly coastal measurements using Python and CronJob scheduling, reducing manual staff effort by ~10 hours/week and improving measurement accuracy by 25%.
Experience
Applied AI & Automation Intern
- Architected and deployed a Hybrid RAG matchmaking system (Prime Ready) replacing 5-6 weeks of manual discovery with instant query responses
- Co-led end-to-end development from concept to production in 2 months with a 4-person team — adopted by 50+ SMBs and NASA
- Built data ingestion pipelines for supplier profiles, iterating on retrieval quality through embedding model evaluation, chunk-size tuning, and hybrid vector search scoring
Data Science Analyst Intern
- Engineered an automated data-processing pipeline for weekly coastal measurements using Python and CronJob scheduling, saving ~10 hours/week
- Designed data validation protocols delivering a 25% improvement in measurement accuracy
Member, Reinforcement Learning Research Lab
- Participate in weekly seminar involving peer review and presentation of current RL research literature
Education
B.S. in Data Analytics & Data Science
Relevant Coursework: Algorithms, Machine Learning, Linear Algebra, NLP, Reinforcement Learning, Information Retrieval, Probability & Statistics, Game Theory