Quillon Lands $1.5M Seed Round for Audit-Grade AI Platform

Quillon, an artificial intelligence startup that helps accountants produce audit-ready technical memos and financial reports, has closed a $1.5 million pre-seed funding round led by Munich-based venture firm 42CAP. Angel investors affiliated with NVIDIA and Roblox also participated, backing the company ahead of its institutional raise. The Sofia and San Francisco-based startup, which recently rebranded from Acclara AI, says its platform is already used by more than 500 accountants across public companies, audit firms, and consulting organizations in the United States.

The company was founded in 2023 by Nikolay Dakov, Ivaylo Stefanov, and Atanas Dobrev. Dakov, who serves as CEO, studied at the University of Oxford. Stefanov, a UCL graduate, serves as co-founder. Dobrev, who holds the title of COO, previously worked as a Senior/Lead Software Engineer at Sporty Group. Together, the three co-founders identified a problem that the broader AI industry had largely ignored: the accounting profession does not just need faster outputs, it needs outputs that can survive an audit.

Technical accounting is the specialized discipline at the heart of financial reporting for public companies. When a corporation completes a merger, restructures its debt, recognizes revenue from a complex contract, or navigates a lease agreement, its accounting team must produce a formal memo that documents exactly how the relevant standards apply to that transaction and how each conclusion was reached. These memos are reviewed by external auditors and, in some cases, by the SEC. Errors or unsupported claims can trigger restatements, regulatory action, and significant financial consequences. It is painstaking, high-stakes work, and the pool of qualified professionals capable of doing it has not kept pace with demand.

This is the problem Quillon was built to solve. The platform is a purpose-built workspace where accountants can research complex standards questions, analyze contracts, benchmark against peer company filings, and draft complete technical memos in a single integrated environment. At its core sits a proprietary knowledge graph of accounting standards and connects them to real-world filings through a direct integration with EDGAR, the SEC’s electronic database. When an accountant poses a question, the system walks through the analysis step by step, linking every conclusion to the exact paragraph in the relevant standard that supports it. The accountant reviews, edits, and validates each stage, maintaining human oversight throughout.

That traceability is what separates the platform from general-purpose AI tools. Large language models can summarize documents and generate text quickly, but they routinely produce unverifiable citations, fabricate references, and fail to maintain clear links between their outputs and authoritative source material. In an industry where an auditor must be able to defend every line of a memo, that unreliability is a dealbreaker. Quillon eliminates the problem by design, ensuring that every claim in a finished document can be traced to its originating source. The accountant drives the analysis. The AI handles the research burden and the drafting. The audit trail is built in from the start.

Dakov has framed the company’s mission in blunt terms. He has said that for two years his team watched technical accountants try to use general-purpose AI to manage workloads that had outpaced hiring, and that they consistently failed because no auditor will defend output that the AI cannot cite. That observation became the founding thesis, and the product was engineered around it.

The $1.5 million in new capital will fund two immediate priorities. First, Quillon is launching a drafting workspace that has been in development for more than a year. The workspace introduces memo authoring, structured decision trees, and an AI agent designed to guide technical accountants through complex analyses. Second, the company is scaling its U.S. go-to-market operation, with plans to expand adoption among public companies and audit firms over the next twelve months.

42CAP, which led the round, is a seed-stage venture firm founded by Thomas Wilke and Alex Meyer, both former founders of eCircle, one of Europe’s largest SaaS companies, which they sold to Teradata in 2012. The firm, now investing out of its fourth fund generation, focuses on early-stage B2B software companies across Europe with global ambitions and typically deploys between €500,000 and €3 million per deal. Quillon fits squarely within that thesis. The participation of angel investors from NVIDIA and Roblox, whose identities have not been publicly disclosed, adds a signal of confidence from operators embedded in the broader AI ecosystem. Their early involvement, which preceded the institutional raise, suggests the product had already demonstrated meaningful traction before the formal fundraise began.

The company currently operates with a team of eight between Sofia and San Francisco. The target customer profile spans technical accounting directors, SEC reporting managers, controllers, and accounting consultants working with complex U.S. GAAP or IFRS questions and audit-sensitive deliverables. The initial product focus is on technical accounting memos, but the roadmap extends into broader financial reporting workflows, including quarterly and annual disclosures. The long-term vision is a platform capable of handling end-to-end accounting workflows and producing finalized deliverables.

The market timing is favorable. Public company accounting teams are under growing pressure from increasingly complex reporting requirements, evolving standards across jurisdictions, and a persistent shortage of qualified professionals. Firms operating internationally must reconcile differences between U.S. GAAP and IFRS, manage currency effects, and satisfy expanding disclosure obligations. The demand for AI systems that deliver automation without sacrificing auditability continues to rise. With the new funding in hand, Quillon aims to evolve from a research-oriented product into a fully production-ready system and broaden its footprint across the American market over the coming year.