Barcelona-based Galtea has raised $3.2 million, in seed funding to expand a platform designed to test and evaluate AI agents before they are deployed in production. The round was led by Munich-based 42CAP and included Mozilla Ventures alongside returning backers JME Ventures, Masia, and ABAC Nest Ventures. The financing underlines growing investor interest in infrastructure software aimed at making enterprise AI systems more reliable, auditable, and easier to scale.
Funding Round
Founded by Jorge Palomar and Baybars Külebi, Galtea was spun out of the Barcelona Supercomputing Center in 2024 after work rooted in the center’s language technologies research. The company had previously secured nearly $870,000 in pre-seed funding, bringing total funding to roughly €3.5 million, or $4.1 million, to date. Galtea said the new capital will support hiring across commercial and engineering roles while funding further product development.
Product and Positioning
Galtea’s central argument is that building an AI agent is no longer the hardest part for enterprises, while validating its behavior at scale remains costly and time-consuming. Its software generates large volumes of use-case-specific scenarios, including adversarial and edge-case interactions, and evaluates systems against structured metrics. Those evaluations are designed to flag weaknesses tied to accuracy, behavior, safety, and security before the software reaches end users.
The company is also widening access to the platform through a self-service offer and free trial, moving beyond its earlier enterprise-focused rollout. On its website, Galtea describes itself as a pre-deployment layer that complements observability tools by surfacing issues before launch instead of after incidents occur in production. That positioning reflects a broader shift in the AI market, where evaluation and validation are increasingly being treated as a distinct software category rather than an extension of standard QA.
Market Context
The problem Galtea is targeting is particularly relevant for businesses trying to move generative AI projects from pilot environments into real operations. The company has said that the shortage of high-quality, affordable testing data remains a core bottleneck, and cited an estimated $13 billion in annual AI testing costs for large companies across Europe and the United States. Coverage of the fundraise also linked the startup’s momentum to rising demand for evidence that autonomous systems will behave predictably under real-world conditions.
That pressure is even stronger in Europe, where the EU AI Act is increasing expectations around safety, documentation, and compliance in higher-risk use cases. Background material from Galtea’s ecosystem and independent reporting describe the company as addressing the need for proof that AI systems can operate accurately and securely outside benchmark tests and product demos. Mozilla Ventures’ involvement adds weight to that narrative because the fund publicly focuses on trustworthy AI and responsible technology.
Early Traction and Expansion
Galtea already lists Telefónica and ABANCA among its customers, according to reporting around the round, and says users have achieved lower validation costs and faster testing cycles through the platform. In one example shared publicly, the company said it generated more than 6,000 scenarios for a customer support agent and identified 2,164 failed evaluations across seven critical vulnerabilities. Public comments from Palomar indicate the business is concentrating on high-stakes sectors including finance, insurance, healthcare, public administration, and education, while also expanding beyond Iberia with a growing focus on the UK.
For investors, the appeal of Galtea appears to rest less on another generative AI application and more on the infrastructure needed to make AI dependable once it leaves the lab. As enterprises roll out agents into customer support, banking, and other sensitive workflows, the ability to simulate risk and document performance is becoming commercially significant. Galtea’s seed round suggests that testing, evaluation, and compliance tooling is emerging as one of the more durable layers of the enterprise AI stack.