Galtea, a provider of an AI agent evaluation platform, has announced a $3.2 million seed funding round to further develop its platform aimed at generating high-quality, tailored testing scenarios for the different generative AI agents deployed in enterprise environments.
The round was led by German fund 42CAP, with participation from US-based Mozilla Ventures, as well as existing Galtea investors including JME Ventures, Masia, and ABAC Nest Ventures.
Access to reliable and affordable test data is one of the biggest obstacles in the AI sector and a key reason why 95% of AI companies never reach production stage. AI testing is estimated to cost companies in Europe and the US $13 billion annually. Galtea aims to remove this barrier by providing reliable, use-case-specific, time-efficient, and cost-effective testing scenarios. The team also announced the launch of its platform under a new self-service plan, which includes access to a two-week free trial.
Jorge Palomar, a former Amazon employee, founded the company alongside Dr. Baybars Külebi. Both identified the need for this solution during their time together at the Barcelona Supercomputing Center. The team has now grown to 12 people, doubling in size compared to the same time last year.
Jorge Palomar, co-founder and CEO of Galtea, said: “Access to sufficient and affordable test data is significantly impacting the speed at which AI teams can deploy their solutions. Many teams cut corners or rely on generic benchmark tests that are not suitable, in order to avoid extremely high costs. Thorough testing and simulation of accuracy, performance, behavior, and security are the only way developers can understand how their platform will perform in a real-world environment. Galtea provides generative AI development teams with the infrastructure needed to overcome adoption barriers and deploy systems reliably, at scale, and efficiently.”
Galtea provides enterprise AI teams with continuous and dynamic testing scenarios, as well as custom metrics to evaluate their agents, enabling teams to determine the best course of action.
The solution is already in use with clients such as Telefónica and ABANCA, as well as other scaleups, corporations, and public institutions.
Galtea has also launched a new self-service product that offers developers free access to the platform, with all features included. This new offering complements its enterprise plans.
Results achieved so far by Galtea’s clients include an average 71% reduction in AI validation costs, delivering an estimated 10x return on investment.
In one specific example, during work with a top-tier institution, a customer service agent failed 2,164 evaluations across 7 critical vulnerabilities—12 times more than had been identified during internal testing. More than 6,000 test scenarios were automatically generated for the agent through Galtea, resulting in an estimated saving of 600 hours of manual work.
Julian von Fischer from lead investor 42CAP stated: “As companies rush to deploy generative AI, the gap between what models are capable of and what companies can trust them to do is widening rapidly. Jorge and Baybars have built the missing quality assurance layer, based on world-class research from the Barcelona Supercomputing Center, and are already demonstrating its effectiveness with customers in high-risk environments. We are excited to support them in defining how companies can manage and scale AI with confidence.”
In a context of growing adoption of AI agents, ensuring their accuracy, performance, behavior, and security is becoming a key challenge for organizations. Galtea addresses this challenge by making AI testing more accessible and efficient, accelerating reliable and scalable adoption of these technologies.
Mohamed Nanabhay, Managing Partner at Mozilla Ventures, added: “As AI systems become more autonomous, the real challenge lies in proving that they work reliably in the real world. Galtea is building the evaluation layer that gives developers the confidence to deploy and scale AI agents.”
Galtea initially raised $870,000 in a pre-seed funding round in 2024, bringing total funding to date to $4.1 million. The team is using the newly raised funds to expand its commercial and engineering capabilities, with the ultimate goal of enabling AI developers worldwide to access fast and rigorous AI testing.
About Galtea
Galtea is an AI evaluation platform that helps enterprises and developers deploy reliable AI agents in production environments. The main barrier to AI adoption is always the same: high-quality, use-case-specific testing scenarios are scarce, expensive, and complex to generate. Without them, evaluation cannot scale. Galtea solves this fundamental problem by automatically generating hundreds of tailored test scenarios in minutes and evaluating AI systems against them to ensure robustness before deployment.
Born in collaboration with the Barcelona Supercomputing Center (BSC), Galtea combines deep research expertise in AI systems with a commercial platform designed for production-scale validation.
About 42CAP
42CAP is a Munich-based venture capital firm investing in early-stage B2B tech companies across Europe with global ambitions. Founded by entrepreneurs who previously built and scaled two of Europe’s largest software companies, 42CAP focuses on data-driven early-stage businesses and brings hands-on operational expertise to the companies it backs. The firm is currently investing from its fourth fund.
About Mozilla Ventures
Mozilla Ventures is an impact venture capital fund partnering with startups and founders dedicated to building responsible technology and advancing trustworthy, open-source AI. Building on Mozilla’s mission to create a healthier internet, the fund invests in companies across three main areas: trustworthy AI, security and privacy, and social impact.
cloud technology axon
Xcalibur Smart Mapping, recognised as the global leader in airborne...
