A New Funding Round for Digital Onboarding

Prelude, a French startup specializing in digital onboarding, has announced the closure of a $20 million Series A funding round. The investment was led by Harry Stebbings' 20VC, marking its first investment in the company. Other investors include existing backers such as Singular, Seedcamp, Deel, and FDJ Ventures, alongside prominent angel investors like Steffen Tjerrild (Synthesia co-founder), Antoine Le Nel (Revolut CMO), and Barney Hussey-Yeo (Cleo founder). With this new capital, Prelude has raised a total of $27 million since its founding in 2023.

Initially focused on assisting companies with user SMS verification, Prelude has rapidly evolved its offering, transforming into a full-stack platform for the entire digital onboarding process. The company's objective is to help businesses verify and protect their users from issues like fraud, ensuring security throughout the customer lifecycle.

The "Hidden Tax" of Onboarding and AI's Response

Prelude aims to solve what it terms the "hidden tax" stemming from ineffective onboarding processes, which manifests as inflated SMS bills and fraud that slips through control systems. The company criticizes legacy providers, arguing that they charge exorbitant rates, rely on outdated dashboards lacking granularity, and offer limited customer support. This context makes digital identity management a significant burden for many organizations.

To address these challenges, Prelude has developed a bundle of discounted, integrated tools, including a verification provider, a fraud vendor, an identity layer, and a device SDK. This combined solution, according to the company, enables customers to save an average of 40 percent on verification costs while also improving user conversion rates. The advent of AI agents for onboarding and advancements in generative AI and fraud tooling have made it significantly easier to impersonate real users, intensifying the need for more sophisticated solutions.

Trust Architecture and New APIs

Matias Berny, co-founder and CEO of Prelude, emphasizes that the "old playbook" is now broken. "CAPTCHAs no longer stop bots, and a single fraud signal won't tell you who's really there," states Berny. "Distinguishing a real user from a fake one has become a business intelligence problem, not a checkbox. The phone number is becoming the strongest anchor we have, and with the Intel API, it carries more trust than any password or one-time code ever did."

Prelude's platform addresses this need by combining telecom data, network signals, and behavioral patterns into a single "trust profile" per user. This approach allows companies to transition from one-time verification to continuous trust decisions. The company has also launched two new products: Auth API, which enables continuous trust checks across the full user lifecycle, and Intel API, which brings real-time intelligence – such as SIM status and number reputation – directly into onboarding flows. This focus on collecting and analyzing sensitive data highlights the importance of data sovereignty and compliance for companies evaluating such solutions.

Future Prospects and Infrastructure Implications

The new funds will be used by Prelude to expand its telecom partnerships globally, invest in machine learning systems, and grow its 50-strong team across engineering, infrastructure, and go-to-market. The investment in "machine learning systems" suggests an expansion of data processing and AI inference capabilities, which may require robust infrastructure.

For CTOs, DevOps leads, and infrastructure architects, choosing AI-powered fraud prevention solutions implies evaluating significant computing requirements. Managing large volumes of telecom data and behavioral patterns to build continuous trust profiles can necessitate careful infrastructure planning, whether cloud-based, hybrid, or self-hosted. The need to ensure data sovereignty and regulatory compliance, especially in regulated sectors like finance, may drive organizations towards on-premise or air-gapped solutions for the deployment of machine learning models and the processing of sensitive data. AI-RADAR offers analytical frameworks on /llm-onpremise to evaluate the trade-offs between different deployment architectures in similar contexts.