Paraglide joins CICM as agentic AI reshapes credit collections
Paraglide, an agentic AI platform built for accounts receivable (AR), has become a corporate partner of the Chartered Institute of Credit Management (CICM), the UK's sole chartered body for credit professionals. The move follows a $5 million pre-seed round earlier this year, backed by Bessemer Venture Partners, DN Capital and Born Capital, and signals that the credit management profession is now treating agentic AI as an operational reality rather than a distant prospect.
The partnership centres on member education: Paraglide will work alongside CICM to help practitioners understand how to deploy AI agents within their workflows, with an explicit emphasis on human oversight. The framing is notable. Rather than positioning automation as a headcount-reduction lever, both organisations are emphasising what Paraglide's co-founder and CEO Rasmus Areskoug describes as a shift "from doing towards monitoring, managing and approving what agents produce."
Agentic AI enters the credit stack
Paraglide's platform automates the high-volume, low-discretion layer of B2B AR: payment reminders, inbound query handling, billing dispute resolution, and collections outreach. The company says its AI agents engage customers in contextual dialogue and, on average, deliver a 34% reduction in Days Sales Outstanding (DSO), the standard metric for how quickly a business collects revenue after invoicing. That figure is company-reported and unaudited, but the underlying problem it addresses is well-documented: AR backlogs tie up working capital, and manual chasing is both expensive and inconsistent at scale.
The CICM has roughly 10,000 members across the UK and international markets. Bringing an agentic AI vendor in as a corporate partner is a meaningful signal that professional bodies in financial services are moving beyond awareness campaigns into structured capability-building. Sue Chapple, FCICM and Chief Executive of CICM, framed the partnership around practical readiness: "Their focus on education, innovation and practical support will help our members better understand the opportunities and challenges surrounding AI in credit management."
The wider convergence: agentic workflows meet financial infrastructure
The Paraglide-CICM announcement sits within a broader pattern that Disrupts has been tracking: agentic AI is migrating from tech-sector pilot programmes into the operational backbone of financial services infrastructure. Accounts receivable is an instructive beachhead. It is high-volume, rule-governed, and data-rich, making it an early-adoption sweet spot for AI agents that can reason over customer history and communicate autonomously. Once embedded here, the same agent architecture is well positioned to expand into adjacent credit functions: trade credit underwriting, dispute arbitration, and supplier financing decisions.
For the investor community, the pre-seed backers are worth noting. Bessemer Venture Partners has a long track record in enterprise SaaS and fintech infrastructure, while DN Capital has backed European B2B platforms at scale. Their combined presence at the pre-seed stage suggests conviction that agentic AR is a category, not a feature, and that the addressable market extends well beyond UK SMEs into global enterprise treasury functions.
The regulatory dimension adds another layer. The UK's Financial Conduct Authority is still developing its framework for AI decision-making in consumer and commercial credit, and the FCA's focus on explainability and human accountability maps directly onto the human-oversight narrative Paraglide and CICM are jointly advancing. Firms that embed structured governance now, through professional body partnerships and documented oversight protocols, are positioning for a compliance landscape that is still being written. That is as much a strategic differentiator as the DSO reduction itself.
Paraglide was founded by Rasmus Areskoug and Andreas Åström, previously the CFO and Head of Engineering respectively at B2B SaaS company GetAccept, where they encountered the AR problem operationally before setting out to solve it with purpose-built AI agents.