Human rounds
Live matching, scheduled rounds, Spar Night, async replies, and an AI decision after both sides finish.
For investors
Product and vision · July 2026
The argument layer of the internet.
Debatable starts with voice practice: two people, a clock, and an AI judge. The longer thesis is a network where spoken reasoning becomes measurable, portable, and useful.
A training ground for human judgment.
Debatable turns arguments into practice, feedback, and reputation. The longer thesis is infrastructure for proving how people reason when another person pushes back.
Read this first
This is a long-term vision board. It separates the product that exists today from work in progress, long-term ambitions, and exploratory ideas. It is not a product roadmap, financial forecast, or offer to invest.
Solo-builtOne founder, a live production product
15+Debate and professional formats
3 pathsDebate people, debate AI, get judged
Voice firstTimed rounds, not another text box
What exists today
A real product before a grand theory.
Debatable is a working voice-first argument trainer. People can match with another person, debate an AI opponent, submit a round for judging, and receive a written decision.
The wedge is competitive debate, where formats, clocks, roles, evidence rules, and judge expectations are specific enough to punish generic output.
AI sparring
Format-aware opponents across school, university, and professional argument settings, with voice and timed turns.
Judging and feedback
Winner, speaker points, reason for decision, and the next skill to practice.
Proof and identity
Profiles, leaderboards, credentials, communication records, and shareable results.
The thesis
Writing gets cheaper. Reasoning out loud gets more valuable.
AI can polish written work for almost anyone. It cannot make a person answer a real objection, choose under time pressure, or defend a claim without a script. Those moments are increasingly useful signals for schools, teams, employers, and the people themselves.
Practice is fragmented
Finding a partner, a judge, a time, and the right format is still harder than it should be.
Feedback is scarce
Most people can practice once. Few can get a useful decision every time and compare progress across rounds.
Reasoning has no portable record
Essays have transcripts and grades. Spoken argument rarely produces evidence that can travel with the speaker.
The flywheel
Every useful round makes the next round easier to start and better to judge.
SpeakA real timed argument
Get judgedA decision and feedback
Build a recordSkill, style, and reputation
Match betterFormat, level, and goals
Return sharperA harder next round
Long-term vision board
Four horizons. Different levels of certainty.
Live now
Building
Long term
Exploratory
Product today
Argument practice
Human and AI opponents, real clocks, and format-specific structure.
AI decisions
A consistent first-pass judge for every completed round.
Public proof
Scores, rankings, credentials, and shareable results.
Network building
Async debate graph
Claims stay open so people can answer on their own time.
Reliable matching
More context about level, format, availability, and preferred judge lens.
Community rhythm
Scheduled events, repeat opponents, teams, and reasons to return.
Long-term platform
Portable verbal credential
A trusted record of reasoning under pressure, not a polished writing sample.
Argument intelligence
A structured map of claims, objections, responses, and what persuades different audiences.
Judging infrastructure
Transparent evaluation tools for schools, competitions, and professional practice.
Exploration
Audience forecasting
Play-point markets and audience calls that make watching a round participatory.
Reasoning data products
Carefully governed research and model-evaluation uses for structured argument data.
Professional rooms
Courtroom, negotiation, pitch defense, oral exams, and high-stakes rehearsal.
Possible business model
One engine, several buyers.
These are product directions, not financial projections. The current beta is still testing willingness to pay, usage limits, and which buyer feels the problem most sharply.
Practice and proof
More rounds, stronger voices, saved history, and a portable credential.
Practice capacity
More students speaking, consistent feedback, and visibility into progress.
Judging and evaluation
Tools for programs and carefully governed access to structured argument data.
Open questions
The risks are part of the thesis.
Can AI judging earn trust?
Only if decisions are specific, contestable, format-aware, and clearly separated from human tournament authority. The product must show its reasoning, not hide behind a score.
Does the network fill quickly enough?
Live matching is a density problem. Async rounds, scheduled events, AI fallback, and narrow format communities reduce the need for everyone to arrive at once.
Is this only for competitive debaters?
Competitive debate is the wedge because expectations are exact. Courtroom, negotiation, pitch defense, and oral exams test whether the same engine travels.
What should remain private?
Voice, identity, minors, and training data need strict controls. A useful reasoning record cannot come at the cost of participant consent.
Interested in the wedge, the network, or the long-term argument layer?
Talk directly with the founder about what is working, what is still uncertain, and which parts of this vision should remain a vision.