Independent developer. Most of my work is browser-side file tooling and the small applications around it: things that do one task properly, run where the user already is, and do not ask for an upload to do it.
A catalogue of file tools that run entirely in the visitor's browser. There is no upload endpoint on the site: files are read, converted and written inside the tab, which is both the privacy claim and the architecture. PDF, image, audio, SVG and document conversion, each with its own page rather than one generic uploader.
Built and shipped solo: the converters themselves, the WebAssembly and worker plumbing that keeps large files off the main thread, the search and page generation, a typed tool surface for AI agents over MCP, and a public request board on D1 where anyone can ask for a tool that does not exist yet.
Other work is under client agreements and is not published here. Happy to walk through any of it on a call.
A single-purpose web tool where the generic product is doing it badly. Usually browser-side, usually because the files should not leave the machine.
Dashboards, internal tools, catalogue and content sites. Deployed on edge infrastructure, handed over with tests and a runbook rather than a demo.
A build that stalled, an integration that half works, a site nobody can safely deploy. Diagnosis first, written up, then a fix you can check.
Rates depend on scope and timeline. Ask and you will get a number, not a range.
Describe the job rather than the tool you think you need. What the file is, what has to happen to it, and what you do with the result afterwards.