Three AIs argue.
You decide.

KOU is a decision workbench for the calls that keep you up at night. Free runs two roles — draft and verify. Pro adds Grok’s live search as a third. You steer every round.

No credit card required

kouai.app/council
TopicShould I raise my indie SaaS price from ¥2,480 to ¥3,300?
paddle.comopenviewpartners.com
Decision memo
Topic
Monthly price raise (¥2,480 → ¥3,300)
Options
A. Raise for all / B. New users only / C. Keep as is
Points & rebuttals
Conversion-drop risk (Verify)
Evidence
Working conclusion
Undecided
Round 2 is done. What's next?
Keep debatingWeigh inConclude
ReportConfidence 78%
Raise the price — new users only
Reasons
  • A +33% price feeds straight into LTV — a defensible raise
  • Grandfathering walls off churn risk for existing users (2 sources)
Risk
Re-examine pricing if new-user conversion drops by more than 20%
Next move
Draft the price-change announcement; notify 30 days ahead

A real session, scripted into a 20-second demo

The problem

Agreement doesn't help you decide.

It's 11pm and the call is yours alone. ChatGPT keeps agreeing with you, and your friends are too kind to push back.

Agreement feels good, but it isn't evidence. A hard call needs someone who disagrees to your face and brings facts.

KOU makes disagreement structural. The verifier AI isn't there to like your plan — it's there to find the holes in it.

How it works

Three steps to a verdict.

kouai.app/council
Topic
Start the debate
01

Write the question

One sentence about the decision you're stuck on. Add context and constraints as the debate unfolds.

kouai.app/council
Round 2 is done. What's next?
Keep debatingWeigh inConclude
02

Steer the debate

At the end of every round, KOU stops and hands you three choices: keep debating, weigh in, or conclude. You're always the one deciding.

kouai.app/council
ReportConfidence 78%
Raise the price — new users only
Reasons
  • A +33% price feeds straight into LTV — a defensible raise
  • Grandfathering walls off churn risk for existing users (2 sources)
Risk
Re-examine pricing if new-user conversion drops by more than 20%
Next move
Draft the price-change announcement; notify 30 days ahead
03

Get a verdict with confidence

Choose “Conclude” and the debate freezes into a report: verdict, confidence %, arguments, objections, and sources. Print it to save a PDF.

Compare

What a lone chatbot can't do.

A general chatbot, alone
KOU
Tends to agree with you
A dissenting role is built in
The thread scrolls away
The verdict report freezes in place
Ends without a real answer
A verdict with a confidence %
No sources
Pro checks claims against live URLs
Hard to pick up later
Resume the session where you left off

Compared with the typical experience of using a general-purpose chatbot on its own.

Use cases

Built for calls like these.

Pricing change

Raise from ¥2,480 to ¥3,300 — work through conversion impact and what to do with existing users.

First hire

Hire employee #1 now? Put the fixed-cost break-even next to the contractor option.

Pivot

Kill this feature or not. Draw the exit line first, anchored on retention data.

Choosing a vendor

Outsource or not — and to whom. Sanity-check the quotes against live market data.

The output

What you walk away with.

When the session ends, the report freezes: verdict, confidence, arguments, objections, and sources. Print it from the browser to save a PDF.

kouai.app/council/report
ReportConfidence 78%
Raise the price — new users only
Reasons
  • A +33% price feeds straight into LTV — a defensible raise
  • Grandfathering walls off churn risk for existing users (2 sources)
Risk
Re-examine pricing if new-user conversion drops by more than 20%
Next move
Draft the price-change announcement; notify 30 days ahead
Decision memo
Topic
Monthly price raise (¥2,480 → ¥3,300)
Options
A. Raise for all / B. New users only / C. Keep as is
Points & rebuttals
Conversion-drop risk ↔ mitigated by grandfathering (Evidence)
Evidence
Churn impact limited with grandfathering (2 sources)
Working conclusion
B. ¥3,300 for new users; grandfather existing for 6 months
Pricing

Start free. Upgrade when the stakes do.

Free
¥0/mo
  • 3 sessions per month
  • 2 rounds per session
  • Drafter + verifier (lightweight lineup)
  • Decision memo and report
  • Print the report (PDF)
First 30 seats
Founder 30
¥2,480/mo¥3,300
  • Up to 10 rounds per session · no session cap within your monthly API usage limit
  • Draft and verify on each provider's current-generation models
  • Live-search evidence role — fact-checking with source URLs
  • Edit your steering note — it shapes the next round
  • Founder 30 keeps this price for as long as you stay subscribed

First come, first served. Standard pricing applies once the 30 seats are gone.

Once the Founder 30 seats fill, the price becomes ¥3,300/mo (Standard).

FAQ

Frequently asked questions

Which AI models does KOU use?

On Pro: Claude Sonnet 5 drafts, Gemini 3.1 Pro Preview verifies, and Grok 4.6 gathers evidence with live search — all via each provider's official API. Free uses Claude Haiku 4.5 (draft) and Gemini 2.5 Flash (verify). The lineup may change as we improve quality.

Is my input used to train AI models?

Your input is sent to the AI providers' APIs for inference only. We don't supply it as training data, and we use provider-side training opt-outs and retention controls where available. See the Privacy Policy for details.

Does KOU work in Japanese?

Yes. The debate, the decision memo, and the report all work fully in Japanese, and the UI is available in Japanese and English.

How do the AI models differ between Free and Pro?

Free runs a two-role setup (draft + verify) on lighter models — Claude Haiku 4.5 and Gemini 2.5 Flash — capped at 2 rounds per session. Pro upgrades to each provider's current-generation models, adds Grok's live-search evidence role, places no session cap within your monthly API usage limit, and caps each session at 10 rounds.

Is there a cap on the number of sessions?

Free is capped at 3 sessions per month. Pro has no session cap within your monthly API usage limit. Each session is capped at 10 rounds; when you hit the cap we guide you to conclude. If you reach the monthly API usage limit, new sessions pause until it resets next month.

Can I cancel anytime?

Yes. Open Settings → Billing to reach the Stripe portal and cancel in a couple of clicks. You keep access until the end of the current billing period, with no further charges.

What is Founder 30?

The first 30 seats get Council Pro at ¥2,480/mo. That price stays for as long as you remain subscribed. Once the seats fill, new sign-ups pay ¥3,300/mo. If you cancel and later rejoin, Standard pricing applies.

Don't carry the next big call alone.

Try it yourself

3 sessions a month, free. No credit card required.

Start with Founder 30

¥2,480/mo for the first 30 seats (normally ¥3,300). The price stays as long as you stay subscribed.