™Division — Research and New Ventures
Xceln Labs
Labs is the front of the pipeline. It explores what is newly possible, builds prototypes fast, and hands the survivors to the rest of the group as products or platform.
01 — What Labs does
Research that has to survive contact with users.
Labs exists to make the group's next decade cheaper. It runs small, time-boxed investigations into new capabilities and turns the promising ones into working software instead of slide decks.
Nothing ships to the public from Labs directly. Its job is to reduce uncertainty so a product team can commit with confidence.
Capability studies
Short, focused investigations into what new models can actually do reliably — published internally as evidence, not opinion.
Prototypes
Working products in weeks, tested with real users. Most are shut down on purpose; the survivors become divisions.
Agent infrastructure
Experimental orchestration, memory and tool patterns that graduate into the shared platform once proven.
Design research
New interaction patterns for AI interfaces that later become part of the group-wide design system.
02 — How work moves
Four stages, and permission to stop at any of them.
- 01
Explore
A question worth answering, staffed for weeks not quarters. Output is a memo and a demo.
- 02
Prototype
A usable product with real users. Success is measured by retention, not applause.
- 03
Graduate
Proven work moves into the shared platform or spins up as a named Xceln product with its own team.
- 04
Retire
Everything else is closed and documented, so the next attempt starts from what was learned.
03 — How it connects to the products
Labs starts them, Metisen sustains them.
- Xceln LearnBegan as a Labs prototype testing whether an agent could plan a full curriculum from one sentence.
- Xceln StudioGrew out of Labs work on controllable generation and brand-consistent output.
- MetisenLabs hands proven agent and evaluation patterns to Metisen, which hardens and operates them.
- The next six productsEverything in the group's roadmap enters through Labs before it gets a name.
Together the two arms form the group's loop: Labs finds what is possible, Metisen makes it dependable, the product companies make it useful.