A consumer product, a complex workflow and a tool for other developers. These are my own projects; each began with something I wanted to make work better.
Here are the stories behind them, what went into the work, and ways to try it for yourself.
Product design
bellos
A personal fix, made into a product.
It started with a workout app that kept losing my records. I built a replacement, then developed it into bellos: a training product with its own identity, accounts, sync, onboarding and feedback. The work reaches from the rules underneath a workout to the experience of using it at the gym.
Take a look: a focused daily workout, backed by the planning and record-keeping underneath.
Private beta
Meet the product, try the sample app, or join the private-beta waitlist. Free during the beta; works on phone and computer.
My household’s spreadsheet had quietly become its operating system. I turned it into an app without losing its history: imports with a preview and control totals, figures that lead back to their entries, edits that can be undone, and data that still exports to Excel.
Try it: explore a fictional household and follow a total back to the records behind it.
Public preview
A household tool, with a fictional-family browser demo. The Apple Silicon download is unsigned; independent security review is still outstanding.
While training as a software engineer at Codesmith, I helped build KafkaSocks, an open-source project released through OS Labs. We wanted to make it easier to bring live data from Kafka into a web app. I returned to that idea with StreamOtter, carrying it much further.
Getting data onto a screen was the beginning. I wanted the screen to stay honest when something went wrong. StreamOtter makes it clear when information is out of date and handles recovery when the connection returns. I built the developer tools, documentation and interactive river demo around it, so other developers can understand and use what’s underneath.
Try it: drop the demo’s connection, then restore it and watch the data recover.
Open source · release candidate
Published on npm under MIT. The public demo uses a fictional creek and simulated observations with a real data pipeline. The release is still evolving.
A few sentences is plenty: what the business does, what’s dragging or what you’d like to build, and when you’re usually free to talk. Nothing to prepare.
