ALPHA KIT
Pilot protocol
5–10 teams, eight weeks, six exploratory hypotheses, and a public report that includes the failures. Free and voluntary; teams can leave at any time.
To join, email hello@sliceframework.org with your team size, what you are building and which AI tools you use. You get a reply within a week.
Cohort
| Segment | Target | Why |
|---|---|---|
| Solo founders building with AI | 2–4 | The design case for SLICE |
| AI-first teams of 2–10 | 2–4 | Tests decision rights and shared constraints |
| Conventional engineering teams up to 25 | 1–2 | Tests whether SLICE holds outside AI-native habits |
Timeline
| Week | What happens |
|---|---|
| −1 | Confirm participation by email and agree on start date |
| 0 | 30-minute onboarding; Quick Start on real work; collect 2–4 weeks of previous delivery data if it exists |
| 1–8 | Run SLICE; 20-minute weekly flow check; 5-question weekly form |
| 9 | Exit survey and 30-minute interview |
| 11 | Alpha report published, including failures, missing data and recommended changes for v0.4 |
What is measured
The six mandatory metrics from the spec (accepted throughput, cycle time, decision latency, wait at the constraint, outcome resolution, calibration), plus overhead: minutes and fields per Mode 1 slice card. Wherever possible these come from issue-tracker and Git timestamps, never from timesheets. F3 forecasting is off.
Weekly form (5 questions, under 5 minutes)
- Roughly how many minutes went to coordination this week?
- Which decision took longest, and what was it waiting for?
- Where did work wait longest: review, QA, a decision, an approval, or elsewhere?
- Did anyone need SLICE itself explained? For how long?
- One thing you would change in SLICE.
Hypotheses
H1–H6 from the spec, all exploratory: with 5–10 teams and no control group, results show direction, not cause.
How your data is handled
- What we collect: the metrics above, weekly form answers and interview notes. Never source code, customer data or credentials.
- Who sees it: only Milber Ferreira, who runs the pilot. Teams are pseudonymous in all analysis, and nothing that identifies a team or a person is published without written consent.
- Where it is kept: We use Zoho Mail for email. Zoho Mail stores data in the EU. Weekly form answers and interview notes are kept until six months after the alpha report is published, then deleted.
- Data controller: Milber Ferreira, hello@sliceframework.org.
- Legal basis: your consent (GDPR article 6(1)(a)).
- Pilot results are published only in aggregated, anonymous form.
- Your choice: participation is voluntary and based on your consent. You can stop at any time and ask for your data to be deleted: email hello@sliceframework.org.