Paraglide Builds AI to Tackle 40% Late Invoice Rate

0
AI software by Paraglide automating accounts receivable and reducing late invoices

Swedish startup raises $5M to let AI chase unpaid invoices.

Paraglide, a Swedish startup that uses artificial intelligence agents to automate accounts receivable, has raised $5 million in seed funding as it bets that the conversational grunt work of getting invoices paid is ripe for automation. The round, equivalent to roughly €4.2 million, was co-led by Bessemer Venture Partners and DN Capital, with participation from Born Capital and The Nordic Web Ventures. The capital will fund product development and a push into broader European markets from the company’s headquarters in Malmö.

The problem the company is chasing is not obscure. More than half of all B2B invoices in the United States are currently overdue, according to industry data, and roughly 40% of businesses say they lack the time to follow up on unpaid bills. The average American company loses nearly $40,000 a year to the accumulated costs of managing late accounts. For small and midsize firms, which make up 99.9% of all U.S. businesses, the drag on cash flow can be existential. Finance teams spend an estimated 15 days per year simply chasing payments, and the invoices that slip past 90 days overdue rarely get collected at all.

Paraglide was founded by Rasmus Areskoug and Andreas Åström, who previously served as Chief Financial Officer and head of engineering, respectively, at GetAccept, a Y Combinator-backed B2B sales platform that raised more than $30 million across its Series A and B rounds. Areskoug’s background includes stints at Google, The Boston Consulting Group and the venture capital firm TCV, experience that gave him a cross-functional lens on where operational bottlenecks quietly erode margins. Åström spent years building the technical infrastructure that powered GetAccept’s document and deal-tracking workflows, experience that now informs how Paraglide’s agents interact with enterprise financial systems. The pair studied together at Lund University’s Faculty of Engineering before their careers diverged and eventually reconverged around accounts receivable.

“As a former CFO, I’ve seen finance teams drowning in billing queries and tirelessly chasing invoices with reminders that customers ignore,” Areskoug said. “AI agents have already transformed customer support by automating high-volume, conversational work. That same shift is now coming to accounts receivable.”

The product represents a departure from what the AR software industry has offered for the past decade. Legacy platforms from established players largely automate workflows and dashboards. They can generate a reminder email or flag an aging invoice, but they cannot read a customer’s reply, interpret a dispute, retrieve a requested purchase order or capture a promise-to-pay date buried in an email thread. Paraglide’s AI agents do all of that. They operate inside a company’s existing finance inbox, conducting two-way conversations in more than 100 languages, pulling documents from connected ERP and billing systems and escalating to a human only when judgment is genuinely required. According to the company’s own data, between 60% and 80% of payment delays stem not from the invoice itself but from the unresolved emails surrounding it.

The incumbents Paraglide is positioning against are not small. HighRadius, the Houston-based AR and treasury management platform, was last valued at $3.1 billion after a $300 million Series C led by Tiger Global in 2021 and counts more than 1,000 enterprise customers including 3M, Unilever and Hershey’s. Billtrust, a New Jersey-founded order-to-cash software provider, was taken private by Swedish investment firm EQT in a $1.7 billion deal completed in late 2022. Paris-based Sidetrade and mid-market players like Versapay and Gaviti round out a crowded field. But Paraglide’s founders argue that none of these platforms were built to manage the conversational layer of receivables, the back-and-forth about missing documents, disputed amounts and payment confirmations that constitutes the real bottleneck.

Early traction suggests the approach resonates. Paraglide counts mid-market and enterprise clients including Choco, the Berlin-based food supply chain platform, Norwegian software firm Ardoq and Swedish sports technology company Spiideo. The startup has also established a strategic partnership with Chargebee, the revenue automation provider. Initial customers reported a 34% reduction in days sales outstanding and reached what the company markets as “financial inbox zero” within their first week on the platform, a claim that warrants scrutiny as the customer base scales. For context, a reduction of just five days in DSO for a company generating $50 million in annual revenue can unlock roughly $685,000 in working capital.

Alex Ferrara, a partner at Bessemer, framed the investment in characteristically direct terms. “There’s a saying in business that cash is king, and if that’s the case then Paraglide’s ability to free up cash flow makes them an AI kingmaker for finance teams,” Ferrara said. “Accounts receivable is an obvious area where generative AI can drive massive efficiencies.” Thomas Rubens at DN Capital described 2026 as the year of the “System of Action,” where AI agents move beyond observation to execute complex workflows with high-level context.

The timing aligns with a broader shift in enterprise software toward what investors are calling agentic AI, systems that do not merely surface insights but take action autonomously. AI captured nearly half of all global venture funding in 2025, with more than $200 billion flowing into the sector. Seed-stage AI startups commanded a 42% valuation premium over their non-AI peers, according to industry data. Finance, which has historically adopted AI more cautiously than functions like engineering or customer support, is now seen as one of the highest-return areas for agentic deployment. Rather than adding another analytics layer to the finance stack, Paraglide builds agents that do the operational work: reading emails, syncing commitments to ERPs in real time and adapting their approach based on each customer’s payment history and communication patterns.

There are risks. Finance departments adopt new technology cautiously, particularly when it involves AI making decisions about customer relationships. Regulatory scrutiny around AI and data protection is intensifying across Europe, where Paraglide plans to expand. And the startup is entering a market where incumbents have deep integrations and long-standing enterprise contracts. The company plans to hire additional engineers, data scientists and product specialists as it scales, building a team that blends financial expertise with machine learning capabilities. Its initial European expansion will target SaaS companies, logistics providers and professional services firms that invoice at high volume across borders.

For Paraglide, $5 million is a modest opening wager against a problem that touches virtually every company that sends an invoice. Payroll, expenses and financial reporting have already been absorbed by automation. Accounts receivable, with its messy human dependencies and tolerance for delay, has resisted. Whether Paraglide’s AI agents can replicate the nuance of a skilled human collector at enterprise scale, across dozens of languages and financial systems, remains the central question. The late payments are not going away. The question is whether the people chasing them finally can.

Leave a Reply

Your email address will not be published. Required fields are marked *