Knottix is the in-house AI engine behind Kreativ Workspace’s merit-scoring and proposal generation — it drafts, a person finalizes. It’s built to remove the blank page, not to replace judgment.
Already running inside Kreativ Workspace today in a limited, human-reviewed capacity — every proposal it drafts is reviewed and finalized by the founder before it counts. A standalone product version is still in development.
Give Kreativ’s own team a faster, more consistent first pass on internal evaluations and proposals, so decisions start from a draft instead of a blank page.
Human-in-the-loop by design: every AI-generated proposal is reviewed and finalized by a person before it counts, so it assists a decision rather than making it alone.
If it holds up inside Kreativ’s own operations, the same model could help other small teams make faster, less-biased internal calls without hiring a data team to build it.
Generates a first-pass proposal and score from the same inputs a reviewer would use.
Surfaces the reasoning behind a score, not just the number.
Every draft routes to a person for review and final sign-off before it’s acted on.
Built on the Claude API rather than a bespoke, harder-to-audit model.
The core loop running inside Kreativ Workspace today.
Team leads and the founder reviewing AI-drafted evaluations today.
Structured internal evaluations for teams too small to build this in-house.
Screens from the Knottix build. Click a tile to view it larger.