Questions, not deadlines
Roadmaps commit to dates; gates demand answers. The systems we deliver hold an organization's money, compliance and reputation, so before any date we ask one thing: can this idea bear real work yet?
Every gate pairs a single question with a defined bar for evidence. If a project falls short, it either waits at that gate or ends. This is what lets our Labs page show each project at its true stage, never one step ahead.
Six gates at a glance
| Gate | What it asks | Evidence needed to pass |
|---|---|---|
| 1 · Concept | Is the problem real and worth a fix? | A plain statement of the problem, tested with the people who face it every day. |
| 2 · Research | What does the existing evidence show? | A study of prior research, data and current tools, finished before coding. |
| 3 · Prototype | Will it work for actual users? | A minimal working build that demonstrates the idea. |
| 4 · Validation | Does it survive real data and real rules? | Recorded results from trials with real users and real data. |
| 5 · Productization | Can people depend on it? | Security, scale, languages and support, added on Anthedon Core. |
| 6 · Product | Will we vouch for it? | Live with a dedicated page, support and clear ownership. |
1. Concept
A project begins with a problem people actually face in daily operations, in one of the sectors we serve. We describe it in plain words and put it in front of those who deal with it. If it proves imagined, or too small for a system, the work ends at this gate.
2. Research
No code is written yet. First we learn what already exists: published research, the data at hand and the tools organizations rely on now. AI projects also get an ethical risk review at this gate.
3. Prototype
Next comes the leanest build that can show the idea works. It is not a mock-up for a presentation; a real person must be able to finish a real task with it.
4. Validation
Here the prototype faces actual data and actual regulations. Most ideas end at this gate, and we count that as a finding rather than a loss: anything that breaks here would have broken later, at a far higher cost to its users.
5. Productization
Projects that pass validation are rebuilt for production on Anthedon Core, the common base under every product: single sign-on and identity, billing, integrations and the audit log. They also gain support, security reviews and all seven languages.
6. Product
Only now is it released to U.S. customers through Anthedon, with its own page, support and a public roadmap.
Every project ends one of two ways: as a product, or closed. None is abandoned halfway.
Where each project stands now
With gates, reporting status honestly is simple.
- EFCP, our compliance platform for ethical finance based on IEAI standards, sits at gate 2: research.
- The agentic platform, which runs AI agents under governance, is at gate 1: concept.
- Anthedon Private AI, models hosted for one organization alone, is also at gate 1: concept.
- GIVING, the nonprofit fundraising platform, has cleared productization and is now in pilot.
What partners can expect
No Labs project is sold as a product until it passes validation, and we set no dates for projects short of that gate. U.S. organizations and researchers can still join early: under a collaboration agreement, partners pilot prototypes and have a say in which ones become products.
