From VC analyst to founder
Baptiste Nassoy had a front-row seat to the startup world as an analyst at Kima Ventures, one of France's most active early-stage funds. He saw hundreds of pitches. But the idea that pulled him out of investing and into founding wasn't glamorous — it was invoice collection, the repetitive, unrewarding task that finance teams dread and nobody wants to do.
In 2025, Nassoy teamed up with Arthur Guerin, who had worked at mobile gaming company Voodoo, and Antoine Grenard, formerly at PwC. The trio's combined background — venture analysis, high-growth tech, and professional services — gave them an unusual lens on the problem. They'd seen from different angles how much time and money businesses lose to late payments: the EU Payment Observatory found that 52% of European companies reported difficulties caused by late payments in 2024, with suppliers waiting an average of 60.3 days and spending 10.2 hours a week chasing customers.
The insight was that collections hadn't been rethought for the AI era. Existing tools either displayed aging balances or automated a first reminder — then left the hard part to humans. Cleavr would build the AI coworker that carried the whole workflow.
The AI collections coworker
Cleavr's platform starts where the invoice leaves the accounting system. It identifies the right contacts, sends follow-ups through email, phone or SMS, interprets replies, tracks payment commitments and escalates disputes — with human intervention reserved for cases that genuinely need it. A promise to pay pauses reminders until the agreed date. A disputed line item takes a different path. A payment that arrives without matching information gets reconciled.
The results, according to the company, are stark: customers reduce their days sales outstanding by an average of 37% within the first few weeks, collect 40% more cash, and recover 80% of the time previously spent chasing payments. Perhaps most tellingly, the company says it hasn't lost a single customer since launch.
The product philosophy matters here. Cleavr connects to existing tools — Pennylane, Sellsy, Sage, SAP, Odoo, NetSuite, Stripe, Qonto — to retrieve invoices, contacts and payments. Invoice creation and accounting stay where they are; Cleavr handles everything from monitoring to collection. No migration, no rip-and-replace. For supported connections, the company says setup takes minutes and the first reminders can go out the next day.
€8M in seven months
Seven months after founding, Cleavr had signed more than 100 companies across logistics, transport, manufacturing, construction and tech. It was commercially active in France, the UK, Germany, Italy, Spain and Belgium, processing several billion euros in invoices annually. That traction — plus zero reported churn — set up a funding story that wrote itself.
In March 2026, Cleavr raised a €1 million pre-seed round. Six months later, in October 2026, it announced an €8 million seed round led by Varsity, with participation from Kima Ventures, Better Angle, Portfolio Ventures, Kerala, Financière Saint James, Station F, Clover and angels connected to companies including Datadog. Total announced funding reached €9 million.
The speed between rounds is the signal. When a company raises its seed six months after pre-seed, it's not because it ran out of money — it's because investors competed to get in. Cleavr planned to use the capital to hire 20 people across sales, product and technology, and to push deeper into European markets. Nassoy was direct about the ambition: to conquer Europe fairly quickly.
Why it worked
Cleavr's playbook is a template for boring-problem startups. First, the problem is brutally quantifiable: every CFO knows their DSO and every finance team knows the hours they waste. When your ROI is 'we cut your collection time by 37%,' you don't need vision decks — you need a demo.
Second, the wedge was integration, not displacement. By plugging into the ERP stack companies already used, Cleavr avoided the enterprise sales cycle that kills young fintechs. The product could prove value in days, which made the land-and-expand motion almost frictionless.
Third, the founders' backgrounds compounded. A former VC analyst who'd seen what gets funded, a high-growth operator from Voodoo, a PwC-trained professional who understood finance workflows from the inside — each founder covered a blind spot the others had. For a product that lives at the intersection of finance, software and operations, that triangulation was the real moat.
The broader lesson: in the AI era, the most valuable startups may not be the ones building the smartest models, but the ones brave enough to automate the work everyone else finds too boring to touch. Collections is repetitive and unrewarding — which is exactly why nobody built for it, and exactly why Cleavr had the field to itself.
There's a final irony worth noting. Nassoy left Kima Ventures to become a founder — and Kima ended up investing in his seed round. The analyst became the deal. For founders coming from the investor side, Cleavr is proof that the best vantage point for spotting a startup idea isn't always a pitch meeting. Sometimes it's the operational drudgery you watched a hundred portfolio companies suffer through, and finally decided to fix yourself.
The playbook
- Pick a painful, measurable problem — Late payments are painful, measurable and universal — 52% of European companies report difficulties. Cleavr picked a problem where the ROI math sells itself.
- Integrate, don't replace — Cleavr plugs into existing ERPs (Sage, SAP, NetSuite, Stripe) rather than asking finance teams to migrate. Meeting customers where they are removed the biggest adoption barrier.
- Raise on momentum, not need — Six months separated the €1M pre-seed and €8M seed. Raising fast on early traction — 100+ customers, zero churn — let Cleavr dictate terms instead of begging for them.
- Make time-to-value days, not months — Deploying within 24 hours and showing DSO reduction within weeks made the product sell itself. Fast time-to-value is the best growth loop in B2B.
- Automate the exception, not just the routine — The AI doesn't just send reminders — it interprets replies, pauses on payment promises, escalates disputes. Handling the messy middle is what separates automation from a mail merge.