The pitch was familiar enough: replace Excel chaos with real-time financial planning. The twist? A database engine built from scratch in Zagreb.
Farseer, a four-year-old financial planning startup that most Silicon Valley investors have never heard of, closed a $7.2 million Series A this week—one of the heftier Croatian tech deals in recent memory. AYMO Ventures, a VC platform that wrapped its first €52 million close just months ago, led the round. SQ Capital, which wrote Farseer's initial $1.5 million seed check back in February 2023, doubled down.
The round, announced February 16, brings total disclosed funding to approximately $8.7 million since founder Matija Nakić and three co-founders incorporated the business in 2020. Not massive by U.S. standards, perhaps. But for a Central European company building enterprise software in a category dominated by well-capitalized American players like Anaplan and Workday Adaptive Planning, it's meaningful runway.
Betting on Rama
What Farseer is really selling—beyond financial planning and analysis software—is a proprietary calculation engine called Rama. Co-founder and CTO Luka Mijatović designed the in-memory database specifically for multidimensional financial modeling, the kind that brings Excel to its knees when CFOs try to model out quarterly scenarios across dozens of business units.
It's a bold technical bet. Most competitors build atop existing database infrastructure; Farseer went the other direction, wagering that owning the calculation layer would unlock performance that off-the-shelf tools can't match. Driver-based planning, real-time forecasting, reporting without the usual spreadsheet lag—Rama handles it, according to Nakić and his team.
Late last year, Farseer layered on what it's calling "Farseer AI," agentic features that let finance teams generate forecasts or query planning data through conversational interfaces. WhatsApp integration included. Whether that's novelty or necessity depends on how quickly AI reshapes FP&A workflows, but the industry is certainly moving in that direction. Datarails, a U.S. competitor, rolled out its own AI insights module just last month.
From Hrvatski Telekom to Western Markets
Farseer's current customer roster reads like a Central and Eastern European corporate directory: JGL Pharma, Violeta, T-Mobile Croatia, Hrvatski Telekom, ZG Holding. Pharma, consumer goods, telecom, retail. Solid logos, if not household names outside the region.
The Series A capital—earmarked for engineering, implementation, and sales headcount—is meant to change that. Nakić wants Western Europe and North America, markets where Pigment (valued at $1 billion-plus after its 2023 Series C) and Anaplan have been operating for years. Farseer reached €1 million in annual recurring revenue by December 2024, according to third-party data. Team size sits somewhere between 16 and 50, depending on which source you trust.
Scaling into mature markets with entrenched incumbents is a different proposition than winning deals in Zagreb or Bucharest. Farseer will need to prove Rama isn't just technically interesting—it has to be demonstrably better, and fast enough to close against competitors with larger sales teams and brand recognition.
The AYMO Factor
AYMO Ventures' involvement signals something about Croatia's maturing startup ecosystem. The fund targets local companies from pre-seed through Series A, a deliberate bet that the region can produce software businesses capable of competing globally. Farseer joins a short list of Croatian startups securing multi-million-dollar rounds, though the country's tech scene remains modest compared to hubs like Berlin or Paris.
For Apertu Capital, the other seed-stage backer alongside SQ Capital, the Series A validates an early thesis. Whether it validates Farseer's approach to enterprise planning is another question—one that will be answered in boardrooms in London, Munich, and maybe eventually San Francisco.
Nakić and his co-founders (Zrinko Dolić and Matej Trbara round out the quartet) are betting that owning the database layer matters. That Rama's architecture can outperform what incumbents have built. That AI-augmented planning tools are table stakes now, not differentiators.
It's a lot to prove with $7 million. But then again, they've already built the hardest part—a database that works. Selling it to CFOs who've never heard of them? That's just the next problem.
