On This Page
Diligent AI builds autonomous AI agents that automate know-your-customer (KYC) and anti-money-laundering (AML) document workflows for fintechs and banks. Founded in 2023 by Edoardo Maschio (ex-BCG, ex-Rocket Internet) and Ahmed Gaber (ex-CTO of Billie), the company operates from London and Berlin.
Overview
Diligent AI is a Y Combinator-backed startup focused on financial crime compliance automation. In March 2026, the company closed a €2.1 million seed round led by Speedinvest and Shapers, with Y Combinator continuing as a backer. Total disclosed funding stands at $3 million, per TechFundingNews. Angel investors include founders and executives from N26, Allica Bank, IDnow, Billie, and Cybersource (acquired by Visa). That roster represents operators who have run compliance functions at scale in regulated environments, providing practitioner-level product validation rather than purely financial backing.
The platform targets the operational bottleneck where compliance teams spend most of their time: routine level-one (L1) tasks such as reviewing registry extracts, verifying proofs of address, investigating legal entity risk, and remediating false positive alerts from AML screening systems. Diligent's approach uses large language models (LLMs) to follow a customer's existing workflows and policies, producing a comprehensive audit trail for each decision. The company explicitly contrasts this with earlier AML automation approaches that generated opaque pattern-matching rules. As CEO Edoardo Maschio stated in March 2026: "When you strip away repetitive tasks, you free up the human mind to focus on judgment and strategy. It's decision-making instead of data processing."
Named production customers confirmed across independent press sources include Flywire (NASDAQ: FLYW), Allica Bank, Alma, Teya, Tamara, Scalapay, and Digital Garage (TYO: 4819), with deployments spanning Europe, the Middle East, the US, and Japan. The most specific outcome data comes from Scalapay, whose Head of Operations Gabriele Alessi reported a 65% reduction in manually reviewed risk queues and 6,000 hours saved annually from manual reviews within the first year of deployment. Both figures are customer self-reported via Diligent AI and TechFundingNews; no independent third-party verification exists.
Speedinvest fintech investor Julien Lézé framed the investment thesis directly: "As AI drives an exponential rise in the volume and sophistication of fraud, compliance operations cannot scale proportionally. The only viable path forward for financial institutions is to fight fire with fire."
Seed funds will support engineering expansion, new AI agent development for additional AML task types, and broader go-to-market rollout across the UK and Europe. As of March 2026, the company is actively recruiting across engineering, machine learning, and sales functions.
How Diligent AI Processes Documents
Diligent's pipeline centers on two discrete agents rather than a general-purpose document extraction engine. The Customer Due Diligence (CDD) Agent verifies companies worldwide by pulling data from official registries and submitted documents, then produces a structured risk assessment. The Name Screening Agent screens individuals and entities against global sanctions, politically exposed persons (PEP), and adverse media databases, applying AI-powered false positive detection to reduce the alert volume that reaches human reviewers. A third agent covering additional AML operations tasks is in development, with no release date disclosed.
Both agents operate through a RESTful API with sandbox and production environments. Rate limits are set at 10 concurrent CDD cases and 20 concurrent name screening requests per account, reflecting the platform's current scale as an early-stage product rather than a high-throughput batch processor. Integration requires no replacement of existing compliance systems: Diligent connects to current tooling via API or through its own UI, with the company claiming deployment timelines of hours to days.
The LLM-based architecture means the system follows natural-language policy documents and workflow descriptions rather than requiring rule configuration by engineers. Coverage from junto.space notes that explainability and logging are described as core product features rather than optional add-ons, which maps directly to regulatory pressure from EU AML directives and UK FCA expectations around explainability in automated decision-making. This is a vendor positioning claim; independent audit of the explainability architecture has not been published.
Unlike general-purpose intelligent document processing platforms that handle diverse document types across industries, Diligent's scope is deliberately narrow: KYC and AML documents in financial services, processed according to the customer's own compliance policies. This vertical focus places it in direct competition with compliance-specific automation vendors rather than horizontal document AI providers.
Security architecture follows zero-trust and least-privilege principles. Customer data is stored in siloed environments, never used to train or update Diligent's models, and encrypted across storage and networks. The platform holds SOC 2 Type II certification under the AICPA framework, independently audited by third-party specialists.
Use Cases
Fintech merchant onboarding
Payment companies and buy-now-pay-later providers face high volumes of merchant applications requiring due diligence checks before activation. Manual review creates a bottleneck that slows legitimate merchants and adds operational cost. Diligent automates the routine checks, routing only higher-risk or complex cases to human reviewers. Scalapay, which serves 8 million customers and 20,000 merchants, achieved a 65% reduction in manually reviewed risk queues and saved over 6,000 operational hours annually. Alma used the platform to scale merchant onboarding without adding headcount, with Head of Global Operations Maureen Gorman noting that Diligent "eliminated repetitive tasks for our team, supporting the quality of our processes by making them consistent and auditable."
AML alert remediation
AML screening systems generate large volumes of false positive alerts against sanctions and PEP lists. Compliance teams spend significant time investigating alerts that turn out to be name-match coincidences rather than genuine risks. Diligent's Name Screening Agent automatically investigates each alert and produces a clear, auditable recommendation, reducing alert-handling time and improving decision consistency. Tamara's Fraud Strategy Manager Karan Khanna reported that the agents "helped us build a stronger and more systematic approach to risk detection." The company launched this agent as a dedicated product following demand from existing customers managing high alert volumes.
KYC document review
Banks and regulated fintechs must review identity documents, corporate registry extracts, and proofs of address as part of customer onboarding and periodic review cycles. Diligent's CDD Agent handles this document review layer, cross-referencing submitted documents against official registry data and flagging discrepancies for human escalation. The KYC document verification workflow is the platform's foundational use case, with deployments confirmed at Flywire, Allica Bank, and Teya across multiple geographies.
Technical Specifications
| Feature | Specification |
|---|---|
| Document types | Corporate registry extracts, proofs of address, KYC identity documents, AML screening alerts |
| Input formats | Not publicly specified |
| Output formats | Structured risk assessments, audit trail reports, screening recommendations |
| Processing pipeline | LLM-based agents following customer-defined workflows and policies |
| API/integration | RESTful API, sandbox and production environments, webhook support |
| Deployment options | Cloud (API-first), UI available |
| Certifications | SOC 2 Type II |
| Claimed accuracy | Not publicly disclosed; 65% manual review reduction reported by Scalapay (customer self-reported, unverified independently) |
| Rate limits | 10 concurrent CDD cases; 20 concurrent name screening requests per account |
| Founded | 2023 |
| Total funding | $3M (€2.1M seed closed March 2026, led by Speedinvest and Shapers) |
Resources
- Diligent AI website
- API documentation
- Security overview
- March 2026 seed funding announcement
Company Information
London and Berlin. Founded 2023 by Edoardo Maschio (ex-BCG, ex-Rocket Internet) and Ahmed Gaber (ex-CTO of Billie). Y Combinator alumni. Total funding of $3 million, including a €2.1 million seed round closed March 2026 led by Speedinvest and Shapers. Angel investors include founders and executives from N26, Allica Bank, IDnow, Billie, and Cybersource. SOC 2 Type II certified. Employee count not publicly disclosed. Actively hiring across engineering, machine learning, and sales as of March 2026.
For competitive context: Diligent's $3 million seed-stage raise is small relative to category peers. EU-Startups noted that Resistant AI raised €21 million Series B and Hawk raised €51.8 million Series C in the same period. Diligent's bet is that vertical depth and analyst-focused design beats horizontal scale at the seed stage. Whether that holds as the company moves upmarket toward larger banks will depend on whether its agent architecture can handle the volume and complexity requirements of tier-one institutions, a question the current evidence does not answer.