Hi all,
I’m working on a privacy-focused AI project and could use a hand from someone who knows the local-inference space well.
The idea is simple: a desktop application (Windows and Mac) that runs a capable open-source model fully offline on the user’s own machine. No cloud, no account, no telemetry. The key thing is that it has to be genuinely plug and play, the user opens it and it just works, with nothing to install, configure, or choose.
Ideally we’d build our own custom solution rather than just wrapping an existing runtime like llama.cpp or Ollama. We want something that’s genuinely ours, purpose-built and self-contained, so we’re open to the right person helping us design it from the ground up rather than assembling off-the-shelf parts.
Where I’d love input or help:
It’s a paid engagement, delivered as open source with the code handed over on completion. Happy to sort out terms properly.
If you’ve built something like this, or have strong opinions on how you’d approach it, I’d really like to hear from you. Drop a comment or send me a message. Thanks!