Why we build Visiyon.

How Visiyon started

Visiyon began as a self-hosted chat interface for large language models, built because the alternatives at the time were either closed, expensive per-seat SaaS tools, or open-source projects that were difficult to run reliably outside of a single developer's laptop. The goal was straightforward: take the open-source ecosystem around LLM interfaces, harden it for real deployments, and keep the core genuinely open rather than open-in-name-only with the useful parts locked behind a paywall.

That starting point shaped how we still build today. Every Visiyon platform has a public, inspectable core. Anyone can read the code, run it on their own servers, and see exactly what happens to their data — because the same code that runs on our cloud is the code you can self-host.

What we actually build

Visiyon isn't a single product — it's a set of platforms sharing common infrastructure: authentication, access control, model routing, and a plugin architecture for extending functionality. The first and most mature platform is our AI chat interface, which supports retrieval-augmented conversations, tool/function calling, and per-group model permissions for teams. Additional platforms are being developed on the same foundation, each aimed at a specific workflow rather than trying to be a single do-everything product.

We think this approach — one open core, multiple focused platforms — produces more maintainable software than bolting unrelated features onto a single monolith. It also means improvements to authentication, logging, or access control benefit every platform at once, instead of being reimplemented per product.

Open source as a default, not a marketing line

We publish the core platform's source code and accept external contributions, bug reports, and feature requests through our public repository. Self-hosting isn't a stripped-down "community edition" — it's the same software we run in our own cloud deployment, and it stays that way deliberately. Enterprise customers pay for deployment help, direct support, and access-control features layered on top, not for a version of the software that outperforms the free one.

This matters to us because we've been on the other side of that trade-off — evaluating vendor tools where the "open source" tier was functionally a demo. We didn't want to build that.

Who we build for

• Individual developers and hobbyists running the open-source core on their own hardware or a cheap VPS.
• Small teams that want a shared AI interface without sending every message to a third-party SaaS vendor.
• Organizations that need SSO, per-group model permissions, and direct support for a production deployment.

How the team works

Visiyon is a small, focused team. We run our own infrastructure, dogfood our own platforms internally, and prioritize stability and clear documentation over shipping every possible feature. Development happens in the open where practical — releases, roadmap items, and discussion threads are visible on our GitHub, and community feedback there directly shapes what gets built next.

Get in touch

Questions about the project, contribution guidelines, or enterprise deployments can be sent through our contact page, or raised directly as an issue on GitHub if it's a technical question about the open-source core.