Henrik Albihn, MS
Applied AI Scientist
Building systems that work quietly. Thinking about the strange loop we're living through.
I build AI systems that serve real needs—not impressive demos. My background is strange: economics, bartending through grad school, now building production ML at scale. Turns out reading people and reading data aren't that different.
I work across modalities—video, audio, text, spatial reasoning. The best solutions come from understanding actual problems, not filtered requirements. Strategic laziness as a design principle.
Experience
Building systems that scale, learning what matters.
Joined as engineer #2 alongside a team from Amazon Alexa, Lyft, AWS, and Elastic. Building mission-critical enterprise AI solutions for Fortune 500 clients.
Built an AI-powered natural language to SQL engine that enables non-technical users to query data at 200x lower cost than traditional analyst workflows.
Led a team of engineers in research, design, and implementation of multimodal AI using real-time sensor data from AR/VR headsets (Quest 3, Vision Pro).
Led team of scientists and engineers building state-of-the-art size recommendation models serving millions of shoppers across 29,000+ brands and 21+ billion products.
Built supervised ML models predicting student course outcomes with >90% accuracy. Spearheaded ETL automation of university information systems.
Education
Formal training, unconventional path
Thesis: "User-Product Recommendations via Matrix Factorization & Stochastic Gradient Descent"
Thesis: "Demand Decomposition in the Market for Plant-Based Meat Substitutes"
Publications
Thinking out loud about AI systems
AI in Plain English
AI in Plain English
Level Up Coding
AI in Plain English
Level Up Coding
AI in Plain English
AI in Plain English
Level Up Coding
AI in Plain English
AI in Plain English
AI in Plain English
Projects
Things I've built. Some work, some teach.