2025 was a year when investors finally accepted that uncertainty isn’t going anywhere, according to our conversations with VCs. What has slowly changed, though, are their priorities. While check sizes and valuations keep rising, the narrative is no longer decisive. Real usage and whether a startup is genuinely hard to copy are becoming the defining factors.
We asked investors from Kibo Ventures, Presto Ventures, and Airbridge Equity Partners to name the startups they believe are worth following in 2026.
Javier Torremocha, co-founder and Managing Partner at Kibo Ventures
Biorce
Biorce is a Barcelona-based company that uses AI to streamline clinical trial execution, helping pharma and biotech teams plan, monitor, and analyze trials more efficiently. What sets the company apart is its ability to reduce operational friction and surface risks earlier in the process, shortening timelines and improving data quality. As a result, customers can bring drugs to market faster while strengthening regulatory readiness and overall trial performance.
German company Ramblr is building the data and tooling layer needed to make AI work reliably in the physical world. Its platform helps companies capture and structure real-world video and operational data so AI systems can learn from actual human work and machine activity. This focus on real-world data gives Ramblr a strong edge as AI expands into areas like robotics, visual inspection, and on-site operations, where context and reliability matter far more than hype.
Abacum is a Spanish scale-up providing a collaborative financial planning and analysis platform that helps finance teams turn data into strategic insights. The company is built around a clear focus on the mid-market, a segment long stuck between spreadsheets and overly rigid enterprise tools. Abacum's flexible FP&A solution for fast-growing teams that need better visibility without added complexity, and its strong traction in the US and consistent wins against legacy players, point to a clear and defensible market position.
HappyRobot, part of the European diaspora in the US, is building AI agents for logistics and supply chain teams to automate high-volume, repetitive workflows such as calls, scheduling, and day-to-day coordination. What sets the company apart is its strong focus on real-world operational use cases, with AI deeply integrated into existing logistics systems. As enterprises increasingly deploy AI in mission-critical environments, HappyRobot is well-positioned to become a key player in operational automation.
Pepijn van den Berg, Investment Manager at Airbridge Equity Partners
The enterprise spend management market is massive, yet remarkably underserved. Most procurement teams still rely on fragmented Excel sheets and expensive consultants to make sense of their data. Mithra is changing that.
The company has built an AI-powered spend categorization platform that automatically cleans, structures, and classifies enterprise spend data, instantly surfacing tangible cost-saving opportunities. What sets Mithra apart is its rare combination of quality, speed, and affordability — three attributes procurement leaders seldom find in a single solution.
Before Mithra, many enterprises couldn’t fully trust their own spend data, let alone extract actionable insights. Now, they can move from messy inputs to clear, defensible savings opportunities in a fraction of the time.
AI is redefining how products are designed and shipped. Teams that still wait weeks for design approval risk falling behind. UXpilot was built for this new reality.
The platform enables product teams to generate professional, on-brand designs in minutes rather than weeks. With integrations across leading design and code-generation tools, plus real-time multi-user collaboration, UXpilot streamlines the entire design-to-build workflow.
Coupled with a smart go-to-market strategy that blends product-led growth with a land-and-expand motion, the company has seen rapid adoption and is well-positioned to accelerate further in 2026.
Technology has made music creation more accessible than ever. The next challenge is discovery: helping artists break through the noise.
With that mission in mind, un:hurd positions itself as the music marketing co-pilot for artists, managers, and labels. Serving more than 200,000 artists globally, the platform uses proprietary algorithms and data from streaming and social channels to measure, plan, and execute high-impact marketing campaigns.
By turning complex marketing decisions into clear, actionable steps, un:hurd helps talented artists find their audience and grow faster than ever before.
Tichomir Jenkut, Partner at Presto Tech Horizons
GRAET is a clear number one for me, led by the powerhouse duo Kroni and Daniela. Kroni is the kind of founder who doesn’t just push through walls — he breaks through mountains. He’s building in a space that has seen very little real innovation, yet one he understands inside out. Given his drive, execution tempo, and deep market insight, I see a very low chance of him not succeeding – and a very real chance of GRAET fundamentally redefining how the sports market operates.
While most AI tools today focus on voice, Matt is bringing his expertise and passion into the broader world of sound. By deconstructing how music is produced, Deep Noise is taking on one of the most complex creative workflows out there. The ambition and technical depth position it well to meaningfully change the music production industry.
SignalPlays
It’s rare to find a founder who understands product as deeply as Samuel. After several startup stints, he decided it was time to build his own company — and SignalPlays reflects that maturity. The company approaches lead generation from first principles, focusing on identifying genuine common ground between business development teams and prospects. That insight translates directly into performance, with results already well above market benchmarks.
E2B is one of those companies where the founder-market fit is immediately obvious. The team is building core infrastructure for AI agents, tackling an unsexy but absolutely critical problem: safe, scalable execution environments. While many teams chase flashy demos, E2B goes back to first principles — if AI agents are going to be real, they need robust sandboxes. The ambition is big, the technical depth is real, and the space remains wide open. I see a clear path for E2B to become foundational plumbing for the entire AI agent ecosystem.
