One commitment ยท One focused review
Tell us the decision. Get a recommendation you can check.
Describe the commitment, what it should achieve and the case for approving it, and add the documents behind it. Decision Review tests the assumptions, the alternatives and the work needed to deliver the result.
It concentrates on what could change the action, the terms, the timing or the acceptable downside. It doesn't recommend another study or pilot unless one could change the answer.
In about a minute
A recommendation, a short plan and the figures recomputed by code
A few minutes later
The in-depth review, audited before it reaches you
At 30 and 90 days
We ask what happened, so later reviews can learn from it
How it works
Four steps, no engagement to set up
You don't need to commission a new business case first. Start with what you have.
Describe the decision
Answer the questions a reviewer would ask: what you'd approve, the result it should produce, what you'd commit, the options you considered and the work it needs. Save a draft and come back.
Add your documents
Optional: the business case, budget, proposal or records, up to 5 MB in PDF, Word, Excel, PowerPoint, text or CSV. You confirm you're authorized to send them.
Run the review
The recommendation arrives in about a minute. The audited in-depth review follows a few minutes later.
Record what happened
Note what you decided. We ask at 30 and 90 days, and later reviews for your company draw on earlier decisions and their outcomes.
What Decision Review examines
The whole case, not just the numbers
Separate approvals can each pass while the expected result still isn't achievable as a whole. Decision Review looks at how the pieces connect.
Evidence and assumptions
What your records establish, what is estimated, and which unresolved assumptions could change the choice. Each figure's inputs are marked as document facts or assumptions.
Alternatives
Other ways to achieve the result, including a smaller commitment, delaying, or continuing with the present approach.
Required work
What must start, change, continue or stop for the benefit to happen, and who owns it.
Operating feasibility
Whether those owners have the authority, skills, information, funding and time; how handoffs, approvals and dependencies affect the plan.
Economics and funding
Costs and benefits recomputed by code, the funding required, and how far each key figure can move before the recommendation flips.
Contradictions and doubts
Figures that disagree between documents, evidence that points the other way, and the objections of people who disagree.
Your other options
You could do this another way
Each of these can be the right call. Each also has a limit worth knowing.
What Decision Review adds: one review that connects the business case, the operating work and the alternatives, in minutes, with every figure checked by code, without repeating the reviews your people have already done.
What you get
The recommendation first, with enough evidence to question it
In about a minute
The quick answer
- The verdict: a one-line headline and one of five findings
- The plan: short advice and the next steps
- The figures, recomputed by code, each input marked as from your documents, with the passage, or as an assumption
- What would change it: how far each key figure can move before the answer flips
- The reasoning: the premises it rests on and what changes if each is wrong
- A document check of your files: figures that disagree, totals that don't add up
A few minutes later
The in-depth review
- A full review of the evidence, alternatives, costs, funding and the work required
- Audited by an AI audit stage and by code checks on its quotes, figures and plan
- Corrected and audited again if it doesn't pass, or held for our staff to review
- Questions answered: if a review asks you something, your answer starts a new run, and the original is kept
- A record of the sources, versions and checks behind it, to print or save as PDF
- Proceed
- Proceed with changes
- Test before committing
- Get specific evidence first
- Do not proceed
An illustrative example
A deadline isn't a business case
A vendor quote is about to expire, the savings look large, and nobody has confirmed what spending would actually fall. This is the kind of decision Decision Review is built for.
All company details, figures and dates are fictional. Not a client submission or result.
Should we approve a $225,000 order-processing system?
The request
We're a 500-person distributor considering a $225,000 order-processing system. The vendor forecasts $140,000 in annual savings, but finance hasn't confirmed how much spending would actually fall. We aren't planning staff reductions. The quote expires October 23, 2026, but our current system remains supported.
The finding
Get specific evidence first
The purchase may be worthwhile. The request alone doesn't establish the savings, the operating feasibility or the full financial commitment.
What could change the decision
Evidence for the benefits and the required work could justify the purchase or a specific change in scope. Establish what quote expiry would actually cost, rather than treating the date as a reason to sign.
Before you start
Good to know
Where your information goes
What you type and the text of your documents are sent to OpenAI to be read. Your company is identified only by a random Company Code; its name, address and phone are never sent.
What is read
The text in your files. Scanned pages, images and charts aren't read.
Specialist questions
Decision Review isn't engineering, legal, tax, cybersecurity or valuation advice. When a decision depends on a specialist determination, it names the question for your qualified people or advisers.
Who decides
Decision Review is a new Emergent Skills method under evaluation. It gives a recommendation; the authorized decision owner keeps the decision.
Start with one decision
What are you about to commit to?
Bring one consequential decision: what you propose to approve, what it should accomplish and when you need to decide.