Developer tools are becoming slower, more dependent on cloud services, and increasingly designed around remote services rather than the local machine.
We believe the alternative is software that is native, local-first, measurably fast, and designed to remain useful even when disconnected from the cloud.
Editors increasingly depend on remote services, hosted models, accounts, subscriptions, telemetry, and proprietary workflows.
AI has accelerated this transition. The computer on your desk is becoming a thin interface to systems running somewhere else, with availability, pricing, and capabilities determined by someone else.
The result may be convenient, but it is also slower, less private, less durable, and more difficult to understand.
Our roadmap is guided by three principles:
01 Build software that stays responsive.
02 Keep developers in control.
03 Make performance measurable.
We believe responsive software changes how people work. Performance is not a feature reserved for high-end hardware or small projects, it should remain predictable as codebases and workloads grow.
AI should extend the developer, not replace them. It should work with your files, your models, and your workflows, giving developers the freedom to choose where their code, data, and intelligence live.
Performance should be demonstrated, not merely claimed. We publish benchmarks, methodology, and engineering research so our results can be reproduced and independently evaluated.
Koi is a native code editor for macOS built around immediate interaction, large files, local AI, and programmable workflows.
It is fast enough to disappear, works without an account, and can be extended in ordinary Python.
Read more →We built Keypress.sh to measure keypress-to-paint latency, publish reproducible benchmarks with raw datasets and methodology, and investigate topics such as editor responsiveness, large-file editing, syntax highlighting, rendering pipelines, and programming language implementation.
We believe engineering claims should be supported by evidence that others can reproduce and verify.
Hackerman is an independent software company building developer tools for the AI era. We believe software should remain responsive, developers should remain in control, and engineering claims should be backed by reproducible evidence.