Timefold tackles the same class of problems, but I thiink their's need a backend while mine runs in browser entirely. So mine works by running the C++ browser by compiling it directly to WebAssembly, so I didn't touch the solver code. The original library has bindings to python, so I did use AI to transpile that interface and tests so that the Typescript API mirrors the Python API.
In somewhat discrete domains with large combinatorial search spaces, I would argue they do. I built this for a conference planner SaaS, which felt very real. Probably have other use cases in routing, dispatching, staff scheduling, conference planning, shift assignment, warehouse picking, bin packing, fleet utilization, job-shop scheduling, matching supply and demand under constraints, etc.
5+ years building robotics and applied AI/ML systems that ship to the real world. Deployed software to a fleet of 30 autonomous golf-ball picking robots, led end-to-end development of an ABB computer vision vision-guided industrial saffron harvester (UNet, NeRFs, Q-Learning), built a hyperspectral drone system for mine detection (UNet, SimCLR), and designed RL-based flight-route planning (PPO, NAF, DQN). Extensive experience connecting perception, sensors, simulation, and robot control all the way through to production. Also built an ATS with an integrated neural reranker capable of processing tens of thousands of candidates (RAG, PyTorch).
English, Swedish. CogSci BSc.
Open to contract/freelance or full-time remote work, including US hours.
Robotics and applied AI engineer focused on getting models to work in the real world. I’ve built autonomous golf ball picking robots, vision-guided robotic systems for saffron harvesting, hyperspectral drone sensing pipelines, RL-based flight route planning, and production ML infrastructure. Open to contract or full-time remote work, including US hours.