Technical Inside is a new company — founded in 2026, in Iași — and one of the first decisions we made was that our tooling layer belongs in the open. Skynet Cluster and TurboProvider are already public, with TurboProvider released under an MIT license. That is not a marketing gesture we backfilled; it is the stance we started from, and it says something specific about the kind of company we intend to be.
The reasoning is straightforward. Infrastructure like local-LLM serving and agent coordination is the sort of thing every serious team ends up needing, and the sort of thing that gets rebuilt badly, in private, over and over. Putting TurboProvider and Skynet Cluster in the open means the friction we removed for ourselves is friction anyone can reuse, and it means the code is exposed to real scrutiny rather than protected by obscurity. Open tooling that people actually run is a stronger claim about competence than any pitch deck.
It also keeps us honest. When your infrastructure is public, you cannot hide sloppy interfaces behind a private repo — the OpenAI-compatible endpoint has to genuinely work, the coordination server has to hold up when someone else points a fleet of agents at it. That external pressure makes the tools better, and it is a large part of why we choose to build this layer in the open rather than treat it as a proprietary moat. The moat is the products we build on top, not the plumbing.
The next step is applied research, published rather than kept internal. Our team's work already sits on research-shaped ground: deterministic PDF redaction with verifiable removal, VRAM-aware model tuning, MCP-native multi-agent coordination. These are areas where a careful, honest write-up — what worked, what did not, what we measured — is worth more to the field than another closed demo, and we intend to publish that way.
That direction is why we are building toward a research partnership with the Faculty of Construction and Installations in Iași. It is an emerging collaboration, not a signed agreement, and we will describe it that way until it is more. But the intent is concrete: pair the applied, ship-it engineering the founding team is built on with academic rigor, and put the results out in public where they can be checked and reused.
Open source and published research are the same commitment pointed in two directions. Both say the work should stand up to inspection by people who did not build it. For a company at the very beginning, that is the reputation we would rather earn — useful things in the open, honestly described — than a wall of claims nobody can verify.