For a doctor in France, the time between a patient's visit and a successful insurance payment is a gauntlet of logins. It involves checking coverage on AmeliPro, booking follow-ups on Doctolib, submitting claims to Almerys, and chasing down rejections. The administrative burden is a well documented source of burnout, a secondary disease state that afflicts the caregiver. Granit, a Paris-based startup founded in 2026, is betting that the solution isn't another piece of software to learn, but an AI agent that already knows the passwords.
The company's core product is a set of autonomous agents designed to operate inside the exact web portals and practice management systems a clinic already uses. By logging in with dedicated credentials, these bots automate the entire billing cycle, from eligibility checks and pre-authorizations to payment follow-ups and reconciliation [PERPLEXITY SONAR PRO BRIEF]. The company claims this approach can reduce administrative time by up to 90%, a figure that, while dramatic and unverified by independent study, points directly at the profound inefficiency it aims to solve [Startup.eu, November 2026].
The Wedge of No Integration
Granit's primary insight is that the biggest barrier to adoption in healthcare is not the cost of new technology, but the operational friction of implementing it. A traditional software solution requires IT integration, staff training, and workflow overhaul. Granit sidesteps this by presenting as a virtual employee that works the existing interfaces. The company says its service can be operational within 48 hours, requiring no complex integration with a clinic's legacy systems [Granit AI, retrieved 2026].
This "no integration" wedge is critical. It allows Granit to target a wide range of healthcare organizations, from small optical businesses to larger clinics, without triggering lengthy procurement or IT security reviews. The agents are described as seizing support directly on platform portals and returning status updates to the human team [getgranit.ai/ressources/almerys-prise-en-charge, retrieved 2026]. For a busy practice, the value proposition is simple, back-office tasks get handled in the background, letting clinical staff focus on patients.
Traction in a Regulated Market
Moving quickly in French healthcare, a system defined by its complex regulations and multiple payer portals, is no small feat. Yet, Granit reports it is already deployed in over 250 healthcare structures across France [Les Echos, November 2026]. This traction, cited across multiple sources in late 2026, suggests the company has found early product-market fit for its non-invasive automation model.
The company's reported growth attracted a $6 million seed round in November 2026, led by Seedcamp with participation from a syndicate including Rerail, Drysdale Ventures, and Bpifrance [Vestbee, November 2026]. The financing is earmarked for expanding across European healthcare markets and adding support for new medical specialties [Startup.eu, November 2026].
| Metric | Value |
|---|---|
| Reported Seed Round (Nov 2026) | 6 M USD |
| Reported French Clinic Deployments (Nov 2026) | 250 clinics |
The Team and the Task Ahead
The founding team presents a mix of technical and operational expertise, though public records show some inconsistency in the exact co-founding lineup. Multiple sources identify Jenny Mansour and Corneille Maréchal as founders, with the company founded in March 2026 [FrenchWeb, November 2026]. Alexandre Sioufi, a developer specializing in AI, is also frequently cited as part of the core team [Spotify, June 2026]. Paul Pietra is noted as building the go-to-market strategy from scratch [Paul Pietra - Granit | LinkedIn, retrieved 2026].
| Role | Name | Note |
|---|---|---|
| Co-Founder | Jenny Mansour | Frequently cited in founding reports [FrenchWeb, November 2026]. |
| Co-Founder | Corneille Maréchal | Listed as co-founder alongside Mansour [FrenchWeb, November 2026]. |
| AI Developer | Alexandre Sioufi | Core team member specializing in AI [Spotify, June 2026]. |
| GTM Lead | Paul Pietra | Building commercial strategy [Paul Pietra - Granit |
Navigating Inherent Risks
For all its early momentum, Granit's path is lined with challenges inherent to automating critical financial workflows in a highly regulated industry. The company's model introduces several specific risks that will test its execution in the next phase.
- Accuracy and liability. An AI agent making an error on a pre-authorization or claim submission could delay patient care or revenue. The company will need to demonstrate near-perfect reliability and have clear protocols for human-in-the-loop oversight and error correction.
- Platform dependency. Granit's agents live at the mercy of the portals they automate, such as AmeliPro and Doctolib. A significant change to one of these interfaces could break functionality, requiring rapid re-engineering. This creates a persistent maintenance burden.
- Scaling trust. Moving from 250 clinics to thousands across Europe means convincing skeptical administrators that an autonomous bot can be trusted with their financial workflows. Granit's answer appears to be its non-invasive setup and rapid time-to-value, but trust at scale is earned through consistent performance over time.
Granit's most plausible answer to these concerns is its reported traction. Hundreds of clinics in a conservative market have ostensibly chosen to run the experiment. The seed funding provides runway to deepen the technology's robustness even as it expands geographically.
The Next Twelve Months
The coming year will be about proving that the French deployment is a replicable blueprint, not a local phenomenon. The seed capital is intended for European expansion, which will involve navigating different national health systems, insurance protocols, and software ecosystems. Success will be measured not just by new country launches, but by sustained usage and renewal rates within its initial French cohort.
A key milestone to watch will be the company's first major partnership or pilot outside of France. Another will be any move toward a more formal integration with major electronic health record providers, which could signal a shift from a pure agent model to a more embedded solution.
The problem Granit is addressing is universal. The administrative burden on healthcare providers is a chronic condition, draining time from patient care and contributing to systemic inefficiency and professional burnout. The standard of care today is a fragmented, manual process. A medical secretary or a doctor themselves must navigate a half-dozen different portals, interpret opaque rejection codes, and spend hours each week on tasks that are purely procedural. Granit is not selling a cure for a clinical disease, but a treatment for this operational pathology. Its bet is that for the exhausted clinician, the best interface is no interface at all.
Sources
- [Vestbee, November 2026] Granit raises $6M to automate administrative tasks for healthcare professionals | https://www.vestbee.com/insights/articles/granit-raises-6-m
- [Startup.eu, November 2026] Paris-based Granit raises €5 million, led by Seedcamp, to automate healthcare administration across Europe | https://www.startup.eu/investments/granit-5-m-seed-11-2026
- [FrenchWeb, November 2026] GRANIT mobilise 5 millions d’euros pour automatiser l’administratif des professionnels de santé | https://www.frenchweb.fr/breaking-news/granit-mobilise-5-millions-d-euros-pour-automatiser-l-administratif-des-professionnels-de-sante/aab2b79d-8f97-4915-b5c4-8d3344e2d15e
- [Les Echos, November 2026] French Tech: ces start-up qui s'attaquent aux lourdeurs administratives des soignants | https://www.lesechos.fr/start-up/deals/french-tech-ces-start-up-qui-sattaquent-aux-lourdeurs-administratives-des-soignants-2261180
- [PERPLEXITY SONAR PRO BRIEF] Web-grounded company brief
- [Granit AI, retrieved 2026] Company website | https://www.getgranit.ai/
- [getgranit.ai, retrieved 2026] Almerys: faire une prise en charge (PEC) optique - Granit | https://www.getgranit.ai/ressources/almerys-prise-en-charge
- [Spotify, June 2026] Interview #4: Alex So yes et le futur du dev avec l'IA | https://open.spotify.com/episode/6Uk9r32q32OGR4tUo6QBjZ
- [Paul Pietra - Granit | LinkedIn, retrieved 2026] LinkedIn Profile | https://www.linkedin.com/in/paulpietra/
- [Nor-eddine Regnard - TOMCAT | LinkedIn, retrieved 2026] LinkedIn Post | https://www.linkedin.com/in/nor-eddine-regnard-8aa77b78/