The Five Peers approach
Every choice deserves
visible reasons.
Five Peers organizes a decision into checkable questions: what you know, what is mandatory, the possible trade-offs and what might change the choice.
You decide.
- 01
Establish the facts
Available for explicit figures
Connect a value to a source passage or your entry. Flag missing information and conflicting amounts. Check and correct the facts.
- 02
Calculate and apply your rules
Available
Compare amounts, lead times and capacity consistently. Mandatory limits come before preferences. Numerical differences are recalculated from the displayed values.
- 03
Assess uncertainty
AI evaluations being developed
TypeSafe will help choose among defined answers, evaluate a criterion against a rubric or assess a statement. Model signals will remain separate from facts and will not represent the likelihood of project success.
- 04
Simulate a change
Available for numerical assumptions
Test a shorter lead time, higher cost or reduced capacity. Compare the hypothetical result with your unchanged starting position.
- 05
Question the available choices
Guided exploration available
Add a renegotiated offer, trial or phased project. Unknown figures still need clarification. Focused checks through TypeSafe will complement these options; they are not active here yet.
A concrete case · fictional data
The right choice depends on your conditions.
Quote A: €12,000, 21 days. Quote B: €9,800, 35 days. Three ways to examine the same decision.
- Your priority is cost.
B saves €2,200 with a lead time 14 days longer. The trade-off is explicit.
- You need delivery within 28 days.
B exceeds that limit. A remains possible, subject to its other terms.
- What if B could deliver in 25 days?
In this simulation B meets the deadline and remains cheaper. The next step is to get the new date confirmed.
TypeSafe evaluates. Five Peers calculates and explains.
TypeSafe is the only AI API planned for this engine. It evaluates focused questions. Five Peers organizes facts, applies rules, calculates differences and tests your scenarios. Explanations are composed from results and prepared wording.
- Select a defined answer · Choice
- For example: which clause answers the question? Include an insufficient-information outcome when relevant.
- Assess a criterion · Score
- For example: is the implementation plan incomplete, partial or detailed? The rubric is visible before evaluation.
- Assess a statement · Noul
- For example: does the passage mention a maintenance warranty? The signal remains a judgment to check against the document.
Today this public experience runs numerical calculations, limits and simulations. TypeSafe evaluations and automatic saving to accounts are being developed.
AI API: TypeSafe · Direct API documentation. Five Peers is a separate product.
Why five approaches?
They address five needs: ground the facts, calculate, represent uncertainty, test a change and explore other options. Use each when the problem needs it; there is no requirement for five model calls.
How do I start simply?
Choose an example or ask your question. Add context or a text PDF, Word .docx, TXT, Markdown or CSV file. Limits: 2 MB and 24,000 extracted characters. Scanned PDFs need a version with recognized text.
What does this experience retain?
The document is processed and deleted. No document is sent to an external AI in this version. Download your comparison to keep a copy. Creating an account does not yet save this trial automatically.
Who makes the final decision?
You do. A priority helps you compare, and a simulation remains an assumption. Quality, warranties and human consequences still need consideration before deciding.