Cognitive engineering for founders
Turn messy business thinking into a structured, inspectable decision — without giving the judgment away to AI.
AI proposes. Founder decides.
Product
Follow one pricing decision from a founder’s first message to a committed record.
01 Think
Bring the real decision, context and uncertainty.
We’re thinking about raising prices 30% this quarter. Margins are thin and our biggest customer is on an annual contract, so I’m not sure we can touch it mid-term. I think churn will stay low — customers love the onboarding, but I don’t actually know churn by cohort. Maybe we only raise for new customers, or add a higher tier instead. If we lose that one account, we’re in real trouble.
Select a phrase to see where it lands on the map.
02 Structure
Turn scattered thinking into alternatives, assumptions, constraints, evidence and risks.
03 Understand
Inspect the structure, challenge weak assumptions and see what still needs validation.
Decision pulse
04 Commit
Make the judgment yourself with the reasoning still visible.
Decision MapExample · a pricing decision
Should we raise prices 30% this quarter?
Raise 30% for all customers
Raise for new customers only
Add a higher tier instead
Gross margin is thin
Largest customer is on an annual contract
Losing the largest account
Churn by cohort
Churn stays low because customers value onboarding
Dashed means unresolved.
Founder authority
Decipath can propose changes to your Decision Map. Nothing enters the map until you apply a proposal.
Your Decision Map
Largest customer is on an annual contract
Proposed by AI
Add · Add node · Proposal only
Add an assumption about how your largest customer reacts to the new price.
You said they’re on an annual contract, but not whether they would stay at a higher price. The decision depends on it.
A proposal is advisory until you apply it. If the map changes underneath it, it is marked superseded.
Provenance
Each card says who put it there: you, arithmetic on your own figures, or AI inference. And whether it still needs validation.
Stated by you
Cash on hand: $1.8M
Monthly burn: $200K
Runway: 9 months
Deterministic derivation
Computed arithmetically from founder-confirmed figures. Not an opinion.
In your words
“customers love the onboarding”
Churn stays low because customers value onboarding
AI-derived
Inferred by structural analysis of your conversation. Challenge before trusting.
Nothing confirmed yet
Churn by cohort
Needs validation
No confirmed evidence yet — treat as an open question, not a fact.
Color is never the only signal. Every origin has its own glyph and its own label.
Commitment
A decision should not become a one-line answer detached from the assumptions, evidence and trade-offs that produced it.
Raise for new customers only
Our biggest customer is locked in for the term. New customers let us test the price without putting that account at risk.
Churn stays low because customers value onboarding
AI analysis · Derived
Medium
New-customer conversion falls sharply after the increase.
Use cases
Different decisions, different structure. The same reasoning system: alternatives, constraints, assumptions and risks, each with its origin.
Hiring
Hire now
Wait a quarter
Change the role
Budget covers one hire
A senior hire ramps up within three months
Time to productivity
Fundraising
Raise now
Extend runway
Grow more first
Runway: 9 months
Running out of cash mid-raise
Investors will price current growth well
Go-to-market
Enter now
Run a test
Partner
Wait
Sales team is two people
Demand in the new market
Founder’s beta
Full access while we learn with a small group of founders.
Start a decisionDecision conversations
Describe the decision in your own words and work through it in conversation.
Structured Decision Maps
Alternatives, constraints, assumptions, evidence and risks, laid out as a map you can inspect.
Founder-controlled editing
Edit the map yourself. AI proposals change nothing until you apply them.
Provenance-aware structure
Every part of the map shows whether it came from you, arithmetic on your figures, or AI.
FAQ
A chat gives you an answer in a stream of text. Decipath turns the conversation into a Decision Map: alternatives, constraints, assumptions, evidence and risks, each marked with where it came from. You can inspect, edit and challenge that structure, and the decision stays yours.
No. Decipath helps you see the structure of the decision. AI can point at weak assumptions and missing evidence, but the judgment is yours, and so is the commitment. When you commit, the reasoning stays attached to it.
Nothing, until you decide. AI can propose changes, such as adding an assumption. A proposal stays advisory until you apply it, edit it or discard it. Anything AI contributes stays marked as AI-originated, so you can always tell it apart from what you stated yourself.
Yes. It is $0 during beta, with full access, while we learn with a small group of founders.
To generate responses, the relevant input is sent to AI providers. Your sign-in cookies, email codes and tokens are never sent to them. Decipath stores only the minimal account, session and decision-memory data needed to run the beta.
During beta, contact the founding team to request deletion of account-associated data. The full detail is on the Privacy page.
Clarity for consequential decisions
Structure the thinking. Keep the judgment yours.
Start a decision