The team behind MLVisor
A small, engineering-led team building MLVisor, VisorEngine and VisorHub at ImperialTechs.
Muhammet Polat
Berkay Başeski
The two co-founders chose to build MLVisor on its own inference engine rather than as a wrapper. That is the slower path in the short term and the defensible one in the long term: we are not bound to any external provider's pricing, rate limits, or roadmap. The team is small and engineering-led — the person writing the engine and the person designing the product sit at the same table.
Built from scratch
The VisorEngine inference architecture was engineered in-house — no external model at its core.
R&D with academia
ImperialTechs runs the research programme with university partners, and the pilots run directly on MLVisor.
One ecosystem
MLVisor, VisorEngine and future products all share the same independent core.
How we work
Engine-First Thinking
We design around an independent inference engine, not someone else's model.
Builder Mentality
We move fast, ship in public, and iterate with our early partners.
Responsibility
We prioritize privacy, security, and keeping customer data in-house.
Collaborative Spirit
We build alongside early partners and the wider AI community.
Want to talk to us?
We are talking to research partners, institutions and investors.