Skip to content
William Peytz

CV

Copenhagen, Denmark

Download PDF

Profile

AI engineer and startup builder pursuing an MSc in Human-Centered AI at DTU. Focused on building real-world AI systems, from adaptive learning platforms to voice-based AI agents handling live customer interactions. Strong in rapid prototyping, full-stack development, and applying machine learning to real user problems.

Experience

  • AI Engineer & Cofounder — Stemma, Copenhagen
    2026 – Present
    • Building Danish-speaking AI phone assistant for handling inbound calls (scheduling, bookings, sales).
    • Designed and deployed a real-time voice pipeline (ASR → LLM → TTS).
    • Integrating telephony APIs and backend systems for live call handling.
    • Running pilot customers to validate product-market fit and quantify missed-call conversion.
  • AI Engineer & Cofounder — Simbolic, Copenhagen
    2025 – 2026
    • Built infrastructure for orchestrating engineering simulations across tools.
    • Explored standardized interfaces for reproducible simulation workflows.
  • Mathematics Tutor — Toptutors, Copenhagen
    2021 – 2022
    • Tutored calculus, algebra, and probability. Developed custom digital exercises with GeoGebra and TI-Nspire.
  • IT & Legal Assistant — Mette Klingsten Law Firm, Copenhagen
    2017 – 2019
    • Automated CRM workflows and document handling. Designed targeted email campaigns.

Selected Projects

Full case studies on the projects page.

  • EloQuiz.dk — Adaptive math learning platform with an Elo-based difficulty system. Full-stack (Vue, Flask, Firebase, OpenAI API), validated through a student pilot.
  • Anomaly Detection — Zero- and few-shot industrial defect detection using DINOv3 on the MVTec dataset. Embedding similarity and patch-level anomaly localization.
  • Audio Classification — CNN/LSTM models classifying 15 instruments. Achieved 96–98% accuracy with tuning and validation.
  • EasyLINC 2.0 — Ride-ordering solution for elderly users with focus on accessibility. Led iterative UX prototyping and user testing.

Education

  • MSc Human-Centered Artificial Intelligence — Technical University of Denmark (DTU)
    2025 – 2027
    Courses: Deep Learning, Computer Vision, Computer Graphics, X-Tech Entrepreneurship.
  • BSc Artificial Intelligence and Data — Technical University of Denmark (DTU)
    2020 – 2025
    Thesis: Adaptive Learning Systems (EloQuiz.dk).
    Relevant: Machine Learning, Reinforcement Learning, MLOps, UX Design, Software Engineering.
  • Economics (1 year) — University of Copenhagen

Skills

  • ML: PyTorch, scikit-learn, deep learning
  • Backend: Python, Flask, FastAPI
  • Frontend: Vue.js, React
  • Infra: Docker, Firebase, Supabase, Vercel
  • AI: LLMs, voice pipelines, OpenAI API

Languages

  • Danish (native)
  • English (fluent)

Get in touch

williamhkp@gmail.com · GitHub · LinkedIn