Compliance officers at more than 100 financial institutions are now running AI agents built by a Paris startup. The agents are designed to do the tedious work: sifting through transaction alerts, screening customers against sanctions lists, and preparing investigation files. For Marble, a three-year-old company founded by alumni of French neobank Shine, that traction has just translated into a €6.5 million ($7 million) Series A [The Next Web, September 2026].
Marble is not selling a black-box AI that promises to replace human judgment. Its wedge is a highly configurable platform that lets risk teams build their own no-code detection rules and then use AI to automate the resulting casework [Fintech Garden, September 2026]. The target is the combined workload of fraud detection and anti-money laundering, a category increasingly called FRAML. The bet is that mid-market banks, fintechs, and crypto exchanges need more flexibility than legacy vendors offer, but lack the engineering resources to build their own systems from scratch.
The FRAML Wedge
Marble’s product surfaces are built around a core workflow familiar to any compliance officer. First, monitoring: users can build rules to flag suspicious transactions without writing code. Second, screening: the platform checks customers against sanctions, politically exposed persons (PEP), and adverse media lists, with updates multiple times a day [GitHub, retrieved 2026]. Third, and most central to its AI pitch, is the investigation suite. Here, AI agents are tasked with gathering relevant data, annotating alerts, and compiling preliminary reports into a unified case manager [The Next Web, September 2026].
The company reports it can be deployed as a hosted SaaS service or installed on a customer’s own servers [Fintech Garden, September 2026]. This flexibility is a pointed feature for regulated entities in finance, where data sovereignty and control are non-negotiable.
Founders from the Front Lines
The founding team, Arnaud Schwartz and Pascal Delange, came directly from the operational pressures they now aim to automate. Schwartz was previously the COO of Shine, the French SME-focused neobank. Delange was Shine’s director of engineering and, before that, its head of data and data protection officer [FinTech Global, September 2026]. Their experience inside a growing fintech gives them a clear read on the pain points of scaling compliance operations with a lean team.
This background likely shaped the product’s emphasis on configurability over magic. They are selling a system that empowers compliance officers, not one that seeks to fully replace them. Schwartz is also an investor in Hexa, a Paris-based startup studio, which became an early backer of Marble [TechCrunch, November 2023].
Traction and the Series A
Customer growth provided the use for the new round. Marble says its platform is in production use at more than 100 institutions, including banks, fintechs, and crypto exchanges, across over 25 countries [FinTech Global, September 2026]. The company does not disclose revenue, but the headcount of regulated customers is a standard proxy for early enterprise traction in this sector.
The €6.5 million Series A was led by Smartfin, a Belgian venture firm focused on sustainable innovation. New investor ADNEXUS joined, alongside existing backers Passion Capital, 42Capital, and Hexa [The Next Web, September 2026]. The round brings Marble’s total disclosed funding to €9 million ($9.75 million) [startup.eu, September 2026]. A valuation was not disclosed.
The funding will be used to expand the AI automation capabilities and grow the team. The competitive set includes point solutions like AMLcheck and iDenfy, as well as broader compliance platforms. Marble’s differentiation rests on its combined FRAML focus and its architecturally open, configurable approach.
Pre-Series A Funding | 2.5 | M EUR
2026 Series A | 6.5 | M EUR
Where the Wheels Could Come Off
The path from 100 customers to becoming a default vendor in regulated finance is steep. The risks are not hypothetical.
- Enterprise sales motion. Landing a pilot with a mid-sized bank is one thing. Securing an enterprise-wide, seven-figure annual contract with a global institution is another. The public record does not yet show a named tier-one bank as a customer, and the renewal motion at that scale remains unproven.
- AI skepticism in compliance. Financial crime units are inherently conservative. While AI can automate routine tasks, final decisions on suspicious activity reports (SARs) carry legal liability. Marble’s positioning,as an assistant rather than a replacement,mitigates this, but convincing cautious compliance chiefs to trust AI-generated case files is a persistent adoption hurdle.
- Competitive depth. The company lists several competitors, but the real competition often comes from in-house builds and the massive, entrenched incumbents like LexisNexis Risk Solutions or Refinitiv. These giants have deep integration footprints and decades of trust, even if their software is less agile.
Marble’s answer to these risks is its wedge: serve the underserved mid-market with a better tool, prove value, and then move upstream. The flexibility of its deployment model and its founder-level understanding of fintech operations are its initial shields.
The Next Twelve Months
The fresh capital from Smartfin and ADNEXUS sets a clear timeline. The next milestones will be measured in product depth and customer tier. Watch for two signals: an expansion of the AI agent capabilities beyond triage into more predictive alert scoring, and the announcement of a flagship customer from the top tier of European banking or payments.
The €9 million in total funding is a serious stake for a compliance automation play. It suggests investors see a path where Marble’s configurable, AI-assisted system becomes the operating layer for a new generation of financial institutions that can’t afford a 50-person compliance team but can’t afford a mistake, either. The question for the board now is whether that wedge is sharp enough to carve out a durable spot next to the industry giants.
Sources
- [The Next Web, September 2026] Marble raises €6.5M Series A led by Smartfin for its open-source fraud and AML platform | https://thenextweb.com/news/marble-raises-6-5m-series-a-smartfin-aml
- [FinTech Global, September 2026] Marble lands €6.5m to automate fraud and AML compliance | https://fintech.global/2026/09/29/marble-lands-e-6-5m-to-automate-fraud-and-aml-compliance/
- [Fintech Garden, September 2026] Article on Marble's FRAML focus and deployment | https://fintechgarden.com/
- [GitHub, retrieved 2026] Marble documentation on screening and monitoring features | https://github.com/checkmarble/docs
- [TechCrunch, November 2023] Paris-based startup studio Hexa raises some funding to launch even more startups | https://techcrunch.com/2023/11/27/paris-based-startup-studio-hexa-raises-some-funding-to-launch-even-more-startups/
- [startup.eu, September 2026] Article noting Marble's total funding at €9 million | https://startup.eu/