Built by Alexandre Guenot.
One conviction: verticalized AI agents beat generic tools on long-term retention. This page sums up who I am, why I'm building Orkelia, and where we're heading.
Solo founder, full-stack builder.
I'm Alexandre Guenot, founder of Orkelia (company based in Antibes, France) since April 2026. A qualified physiotherapist trained in Madrid, it was in the field — during my internships in private practices and clinics — that I saw the problem Orkelia solves: patient requests that come in after hours and go unanswered, where generic tools understand nothing about the business. Rather than put up with it, I chose to solve it myself — and turn it into a product.
I've built 6 vertical branches (Clinic, Restaurant, Transport, Sales, Trading, Shop) because I believe a dedicated AI agent built for a specific business — with its vocabulary, its integrations, its tone — always beats a recycled generic GPT. The proof: 3 live products today, including 1 paid SaaS.
Orkelia is not an agency that subcontracts to OpenAI. It is a system that builds, operates and continuously improves vertical agents for specific industries. Delivery is managed, not self-serve.
The view from the field.
During my internships in private clinics and hospitals, I saw the same problem everywhere: patients who write in the evening, after hours, and have already booked elsewhere before anyone calls them back. No tool addressed this real problem — generic chatbots replied off-topic, outsourced receptionists cost a lot for little value, and booking solutions only handled the appointment.
The need exists across industries: restaurants losing a reservation at 10pm, carriers ignoring quote requests in the evening, aesthetic clinics with 40-60% of Instagram DMs left unanswered. Orkelia is built to solve this precise problem, not to do "AI in general".
May 2026 · raw numbers.
I'm transparent about the numbers: 0 paying clinic client to date. The Clinic product is in Founder Programme (14-day trial no card, 30-day guarantee) with 0 conversions yet. That's partly why I'm building founder-led — every deal counts to validate pricing/positioning hypotheses.
What it's built on.
No-compromise tech infrastructure (eu-west-3, GDPR, DPA provided, electronic signature):
Zero medical data stored. Branded tracking per client domain (CNAME). Total brand isolation per workspace (see the Mynaparrot case with 0% orkelia.com leak on the recipient side).
The goal, set in stone.
By mid-2027: 50 recurring clients spread across the 6 branches, including 5 Enterprise deals (Clinic Group, Tailor bespoke), €100K ARR, a team of 3-4 people (1 ops/customer success + 2 dev). Bootstrap, not fundraised.
No additional branches beyond the current 6 until retention is validated. No pivot toward a generic self-serve tool — managed verticals are the strategy, not a compromise.