I train models, break things, and ship what works.

From fine-tuning video restoration models on A100 HPC to shipping a collaborative AI canvas used in production. I have experience across the whole stack.

My edge is combining ML research depth with product execution and clear communication. I execute quickly, take ownership, and stay reliable under ambiguity and tight deadlines.

Best projects

  • Treyspace logo

    RAG-powered Collaborative canvas where notes, docs, and diagrams are stored as a graph so AI answers can reference spatial information.

    TypeScript | React | Excalidraw | Node.js

  • Bristol Vision Institute logo

    Low-light video denoising research. Currently under review at IEEE ICME as second author. Beating previous SOTA with 29.22 dB PSNR and 0.863 SSIM on BVI-RLV with bidirectional warping and adaptive occlusion masking.

    Python | PyTorch | PTLFlow | RAFT

  • Open-Jet logo

    Agentic Terminal User Interface for NVIDIA Jetson devices for local LLMs without any internet connection. Think Claude Code for edge compute.

    Python | llama.cpp | GGUF quantization | LLMs

  • BECCS Innov logo

    End-to-end venture strategy for a carbon-credit supply chain business — cross-functional team.

    Market research | Financial modelling | User interviews | Prioritisation frameworks

Work Experience

  • Jun-Jul 2025

    Visual Computing Research Intern

    Bristol Vision Institute

    Fine-tuned RAFT/DPFlow on SLURM-scheduled A100 HPC, re-engineered the CNN-RNN recurrent loop for bidirectional warping, and reached 29.22 dB PSNR, 0.863 SSIM, and 0.334 LPIPS on BVI-RLV.

  • Sep 2023-Apr 2025

    Teaching Assistant

    University of Bristol

    Led weekly lab sessions, debugged student code, and graded coursework for up to 200 undergraduates across Linear Algebra, C, Discrete Mathematics, and Data Science.

  • Jul-Sep 2024

    Software Engineer Intern

    Scribblepad Press

    Rebuilt the marketing site, implemented technical SEO and Core Web Vitals fixes, and set up analytics — increased organic traffic by 119%.

Education

  • MEng Computer Science with Innovation

    University of Bristol

    2022-2026

    First Class (current average 81%). Modules: Data Science (92%), Machine Learning (87%), Linear Algebra (82%), Computer Vision (78%).

  • A Levels

    Balcarras Academy

    2020-2022

    4 A*s in CS, Mathematics, Physics, and EPQ. Awards: UKMT Silver five consecutive years, Grey Kangaroo Qualifier.

Essays

My Philosophy

Reports

  • Contrarian Leadership — A case against consensus-driven decisions and for low-consensus, high-conviction leadership.
  • BECCS Supply Chain Report — Commercial feasibility analysis for a carbon-credit supply chain venture, covering policy risk, buyer demand, and unit economics.
  • GCP AI Marketing Strategy — Go-to-market plan for Google Cloud AI services, developed as part of a competitive analysis project.

Get in touch

louisforster64@gmail.com