A bounded pilot
One use case, a written scope, a duration, and success criteria set before the start. A pilot with no exit criteria never ends: it drags on and is judged on impressions.
Propose pilots, interoperability work, events, co-marketing and sector access. An alliance is judged on what it produces — a pilot that runs, interoperability that is proven — and not on the announcement that launches it.
Two families of counterparts, with different expectations. Software vendors and solution providers first: business applications, management tools, data platforms, hardware. The question then is a technical one — what happens when an agent has to read, write or trigger something on your side, and under which versions that holds.
Institutions next: professional federations, chambers of commerce and similar bodies, associations of elected officials, competitiveness clusters, support organisations. Here the question becomes sector access: understanding a sector's real constraints, presenting a subject to members without turning it into a sales pitch, measuring what works and what does not.
The track allows two things, and the fact that there are so few is deliberate:
offer a client a pilot and communicate jointly, on verifiable facts
Anything absent from that list lies outside the track. An alliance does not put deals forward on Blue Lemon Agent's behalf and is not remunerated by the programme in any way: an organisation wishing to recommend Blue Lemon Agent to its members or clients belongs on the referral track, with its own rules.
The track excludes the following:
display a logo without an official source, announce compatibility without versioned evidence and announce an integration without versioned evidence
Compatibility and integration are announced only with evidence tied to named versions, on both sides.
The reason is simple and holds in almost every technical partnership: an integration demonstrated once, in a well-tuned environment, stops working when a version moves on one side or the other. If the announcement came before the evidence, the client discovers the gap in production, and attributes it — often rightly — to whoever did the communicating.
In practice, we ask for three things before publishing anything jointly: the exact versions on both sides, the scope of what was verified, and what was not. The third point is the one people forget, and it is the one that prevents disappointment.
The same requirement applies to logos: a logo is displayed only with an official source, in both directions. We will not display yours without your written agreement, and we ask for the same in return.
One use case, a written scope, a duration, and success criteria set before the start. A pilot with no exit criteria never ends: it drags on and is judged on impressions.
What was verified, on which versions, on what date. That document outlives the people who wrote it, and that is precisely its value.
Workshops, feedback sessions, joint publications. Every piece is reviewed by both sides before release: it is the only way to avoid a capability being described by someone who does not operate it.
An institution opens a door and keeps its independence. It does not recommend a supplier to its members by virtue of the alliance, and it has nothing to gain financially from it.
No. The track's rights are limited to proposing a pilot and communicating jointly; no commission is attached to it, in any direction. An organisation wishing to recommend Blue Lemon Agent and be paid for it applies on the referral track, whose rules are published separately.
When it is proven on named versions, on both sides. This is not a scheduling formality: an announcement made before the evidence rebounds at the first version change.
Yes, with an official source on both sides: a written agreement stating which logo, on which materials and for how long. Without that source, no logo is displayed — displaying one without agreement is exactly what the track excludes.
The answer comes in a sentence we do not alter: If the BLA catalogue does not properly cover your need, BLA can study a custom agent. This possibility remains subject to validation of its architecture, feasibility, compliance and pricing. No lead time, capability or price is announced before that validation, and a pilot is not a commitment to bring something to market.
The programme is open in France, and a translated page does not open it elsewhere: eight checks must be met first — language, support, pricing, connectors, data protection, sector compliance, tax and VAT and capacity to go live. French law, subject to the mandatory rules and overriding provisions applicable locally.
Like the others, through declared states: draft, submitted, under review, awaiting further information, approved, rejected and withdrawn. We ask for a named contact on each side and a first concrete objective — a pilot, an interoperability check, an event. An alliance with no concrete objective remains an intention, and an intention is not something we publish.
Describe the concrete objective you are aiming at first: a bounded pilot is worth more than a general framework.