Galtea has raised USD $3.2 million in seed funding in a round led by 42CAP.
The Barcelona-based company will use the money to develop its platform for generating test scenarios for generative AI agents. Mozilla Ventures joined the round, alongside existing backers JME Ventures, Masia and ABAC Nest Ventures.
Galtea sells software that helps developers and companies test AI agents before deployment. Its system generates use-case-specific scenarios and applies tailored metrics so teams can identify errors, measure performance and assess how systems behave under different conditions.
Galtea argues that testing has become a major obstacle for businesses trying to move AI systems from pilots into live use. It cited an estimate that AI testing costs large companies in Europe and the US about USD $13 billion a year, while broader industry research suggests most enterprise AI projects do not reach production.
The funding announcement coincides with the launch of a new self-service product. Developers can now access the platform directly, while enterprise packages remain available for larger customers.
The business was founded by Jorge Palomar, previously at Amazon, and Baybars Külebi, who holds a doctorate. The pair met while working at the Barcelona Supercomputing Centre, where they identified the need for more targeted testing data for AI systems.
Its workforce now stands at 12 after doubling over the past year. The new funding will support the expansion of its commercial and engineering teams.
Customer base
The platform is already used by organisations including Telefonica and ABANCA, as well as unnamed startups and enterprises. According to Galtea, those customers have seen an average 71 per cent reduction in AI validation costs and an estimated tenfold return on investment.
It also cited one example involving what it described as a Tier 1 institution. In that case, a customer support agent failed 2,164 evaluations across seven critical vulnerabilities, 12 times more than the customer had identified through internal testing. More than 6,000 test scenarios were generated for that agent, saving an estimated 600 hours of manual work.
The issue Galtea is targeting reflects a wider concern in the AI market. Businesses are under pressure to show that AI systems are accurate, safe and reliable before using them in customer service, financial services and other settings where errors can have operational or regulatory consequences.
For banks and telecoms groups in particular, testing has become more important as AI agents take on tasks previously handled by staff or more limited automation tools. That makes it more important to assess not only output quality, but also security weaknesses, problematic behaviour and performance under stress.
"Access to sufficient, affordable testing data is significantly impacting the rate at which AI teams can deploy. Many teams take shortcuts or rely on inadequate generic benchmarks to bypass what are extreme levels of cost. Thorough testing and simulations for accuracy, performance, behaviour, and security are the only way developers can know how their platform will perform in a real-world setting. Galtea is providing generative AI developer teams with infrastructure to overcome barriers to AI adoption and deploy systems reliably, scalably, and efficiently," said Jorge Palomar, Co-Founder and CEO of Galtea.
Investors in the round say they see a growing market for tools that can evaluate AI systems as companies move beyond experimentation. As models become more autonomous, the challenge shifts from building systems to proving they work consistently in real-world settings.
"As enterprises race to deploy generative AI, the gap between what models can do and what companies can trust them to do is widening fast. Jorge and Baybars have built the missing quality assurance layer, rooted in world-class research at the Barcelona Supercomputing Centre, and are already proving it with customers in high-stakes environments. We're thrilled to back them as they define how enterprises govern and scale AI with confidence," said Julian von Fischer of lead investor 42CAP.
ABANCA offered an example of how the software has been used in practice, saying the platform identified issues that had not been found through its own internal checks.
"With Galtea, we uncovered vulnerabilities we would likely have missed otherwise, saved significant engineering time, and improved the reliability of our AI systems. It changed how we approach AI evaluation and governance," said Jorge Romaris, Head of AI at ABANCA.
Mozilla Ventures, which also participated in the round, linked the investment to the wider question of trust in autonomous systems.
"As AI systems become more autonomous, the real constraint is proving they work reliably in the real world. Galtea is building the evaluation layer that gives developers the confidence to deploy AI agents and scale them," said Mohamed Nanabhay, Managing Partner at Mozilla Ventures.
Before this round, Galtea had raised USD $870,000 in pre-seed funding, bringing total funding to USD $4.1 million.