Sharjeel Abbas
Foundertellodb, Brooklyn NY. Building tools for reviewing AI answers and the decisions they inform.
About RigorDesk
An answer should be open to question. That is why we built RigorDesk.
RigorDesk is a workspace for drafting and reviewing proposals, plans, and AI answers. You can write with one model, then ask other models to challenge the work. Their responses, conclusion, and unresolved objections stay together.
We started with a practical question: what did the first answer miss? A second review can surface assumptions or alternatives you had not considered. It can also be wrong, which is why the reasoning must remain available to inspect.
How we approach review.
These principles guide what we build and how we describe it. The useful outcome is a clearer understanding of the work, including its limits.
01
Make the reasoning visible.
A polished answer can hide a weak assumption. We keep model responses, objections, and usage together so you can inspect how a conclusion was reached.
02
Give each review a clear purpose.
Improving a draft and challenging a recommendation are different tasks. Our seven review formats give models different ways to question, defend, and revise an answer.
03
Keep disagreement in view.
An agreement score describes the models’ positions. It does not prove their claims. Read unresolved objections alongside the conclusion and check the evidence that matters.
04
Leave the decision with you.
The result should help you revise a proposal, ask a better question, or spot something to verify. AI reviewers contribute to that work; they do not approve a decision for you.
Who's behind it.
RigorDesk is built by tellodb and was previously called Debate TelloDB. Contact us with product questions, feedback, or partnership enquiries.