Tyler Forstrom

Open to software engineering roles

Software Engineer

Tyler Forstrom

I build well-tested systems that are honest about what they can — and can't — do.

From simulation engines and statistical pipelines to clean web frontends, I turn ambiguous problems into systems that are reproducible, well-tested, and clear. I'd rather ship a result I can defend than a demo that overpromises.

View my work ↓ GitHub LinkedIn Résumé (PDF)
Tyler Forstrom

Selected work

Agent-Based Market Simulator & Strategy Validation Firewall

A from-scratch stock market where the price emerges from thousands of interacting trader-agents — paired with a statistical "firewall" that honestly tests whether a trading strategy has real edge. Across eight strategy families it found none (the correct answer for efficient markets); the one signal that survived is volatility forecasting, competitive with industry-standard models. Pure Python, ~350 tests, tamper-evident result logging.

  • Python
  • Simulation
  • Applied statistics
  • JavaScript
Live demo → Code →

Sketch-to-Image Generation (Deep Learning)

A deep-learning model that turns rough sketches into colorized images. I built a custom U-Net generator and trained it as a GAN (with a PatchGAN discriminator) in PyTorch, using perceptual, edge, and L1 losses plus a multi-stage training curriculum to keep training stable. It's wrapped in a Streamlit app where you can draw or upload a sketch and get a generated image back in real time.

  • Python
  • PyTorch
  • Deep learning
  • GANs
  • Streamlit
Live demo → Code →

Games

Ripple — a wave-interference puzzle

A logic puzzle built on real wave physics: drop emitters and their ripples interfere — reinforcing into bright spots, cancelling into dark ones. Solve each level by placing sources so the pattern lights the right targets and silences the others. Vanilla JS + Canvas; levels are procedurally generated from a known solution, so every puzzle is guaranteed solvable.

  • JavaScript
  • Canvas
  • Wave physics
  • Procedural generation
Play → Code →

About

I'm a computer science student at Cornell, graduating in May 2026, with a minor in business. Most of my projects start the same way: I get curious about how something works and want to see if I can build it myself. I got into investing, so I built a stock market simulator from scratch. I wanted to understand how GANs actually train, so I built a model that turns sketches into images.

What I care about most is whether the thing actually does what I say it does, so I test a lot and try not to oversell my results. I've interned as an app developer at Kahua working on full-stack features, and I helped start an investing club at Cornell. When I'm not coding I'm usually at the gym, training martial arts, gaming, or out on a golf course. Right now I'm looking for software engineering roles where I can build real things and work with people I can learn from.

Tools I work with

Get in touch

I'm open to software engineering roles and interesting problems. The fastest way to reach me:

Email me GitHub LinkedIn