π The Team β Three Founders in Milan
the cell is a three-founder team based in Milan, Italy. We turn ideas into practice by combining AI/full-stack engineering, industrial/hardware design, and brand-led creative strategy.
π€ Idea Evaluation Council
A multi-agent council that debates and scores business ideas. 11 agents organized into 5 teams run a structured debate across 3 rounds: Independent analysis β Cross-examination β Synthesis.
Teams & Agents
π₯ Agent Profiles & Lenses
Each agent brings a specialized perspective to the idea evaluation. The agents below are grouped by the team they belong to and the criterion they own.
π Weighted Opportunity Scorecard
Each idea is scored on 10 criteria, weighted by impact. The final score ranges from 0β100.
| Criterion |
Weight |
Owner(s) |
Definition |
β‘ Debate Structure β 3 Rounds
The council runs a structured debate to reach consensus on an idea's score and recommendation.
πͺ Combined Capabilities Matrix
The team's collective skill set spans AI/ML, full-stack engineering, brand strategy, industrial design, and manufacturing.
| Capability |
Owner(s) |
Maturity / Notes |
π How the Council Evaluates an Idea
- Idea submission: Name, business model, category, and one-line summary.
- Round 1 β Analysis (parallel): Each of 8 analysts reads their skill and writes a preliminary 1β5 score on their owned criterion, citing specific findings and risks.
- Round 2 β Cross-examination (parallel): The skeptic attacks the weakest/most optimistic claims. Each analyst rebuts: "concede what's fair, defend what's solid," then gives a revised score.
- Round 3 β Synthesis: The chair reads the full debate, assigns final weighted sub-scores citing agent evidence, computes the 0β100 total, states confidence, and names the biggest swing factor.
- Output: Full council transcript (all 3 rounds + final scorecard) written to
ideas-generated/{slug}/council-transcript.md.
- Recommendation: advance (pursue further), hold (promising but needs validation), or park (not viable now).