BYOK · Zero server cost · Google Gemini

Let AI agents
tear each other apart,
keep the strategy that survives.

100+ specialized AI agents — advocate, red team, CFO, judge — run a 4-phase debate to produce your market-entry strategy. Not in months, not in hours: in minutes.

Scroll
Benchmark · Food EmpireS$76.2M from CIS8 markets debatedWinner · Philippines 7.455 campaigns ranked$13M playbookROAS 1.8×25,000 stores by M12brand · LaKás KapeBenchmark · Food EmpireS$76.2M from CIS8 markets debatedWinner · Philippines 7.455 campaigns ranked$13M playbookROAS 1.8×25,000 stores by M12brand · LaKás Kape
100+
Specialized Agents
4
Debate Phases
50+
Target Markets
~12 min
To Full Report

How It Works

4 Phases, 100+ Agents

01Intelligence
Historical Benchmark
Agents dissect a benchmark company's financials, market moves, and campaign history across decades.
02Debate
8-Market Battle
Advocate, Red Team, CFO, and Judge agents argue every candidate market at once. The weakest get eliminated.
03Creative
Campaign Tournament
Five campaign concepts are generated, attacked, defended, and ranked in a multi-critic creative tournament.
04Synthesis
Final Playbook
Refinement agents produce the board-ready strategy: brand identity, 12-month plan, budget, and KPIs.

Adjustable Debate Depth

Deeper debate. Same cost.

A shallow debate rushes to a verdict. So depth is a dial (1–10). Each agent runs an N-round attack → rebuttal → counter loop inside a single model call — deeper scrutiny, identical API cost. Same eight briefs, red team at depth 3 vs 10:

Same input · same costDepth 3Depth 10Δ
Markets fully pressure-tested68+2
Total attacks generated4256+33%
Critical findings surfaced3047+57%
Avg. finding detail461 ch540 ch+17%
API calls made110 extra

At depth 10 the red team escalated from generic gaps to second-order flaws — "HIGH modern-trade access is a commercial illusion," "$5M is a test budget, not penetration spend." Deeper reasoning, zero extra calls.

Decision Audit

The winner isn't rubber-stamped.

Every analysis grades its own confidence. Even the top-scoring market ships with a "human review" flag if it still carries unresolved critical objections. Every number comes with its own dissent file.

Phase 2 · Market
Confidence gradeLOW

The winning market is graded HIGH / MEDIUM / LOW. Red-team critical count and CFO status feed the grade.

Phase 3 · Tournament
Open assumptions4 open

The winning campaign clears 7 specialist critics, but unproven assumptions are listed explicitly before any pilot.

Phase 4 · Validation
Unit economicsFLAWED → rewrite

The CFO audits the first draft's unit economics. A FLAWED verdict auto-triggers a full playbook rewrite.

Real Case · CDSG Coffee

A Myanmar conglomerate, entering a coffee market it had never sold in.

Eight countries, one benchmark (Food Empire), zero humans in the loop. The agents came back with a country, a brand, a budget, and a 12-month plan — every claim attacked, every number audited.

Philippines
Winning market · 7.45
LaKás Kape
Generated brand
$13M
Playbook budget
1.8×
Target ROAS
Walk the full case study ↗
🔑

Your key, your cost.

DebateAI never stores your API key. It lives in your browser, and only your browser. Every AI call runs on your own Google Gemini key — a full analysis costs ~$0.10–0.40. Grab a free key at aistudio.google.com.

Get Started →