Artificial intelligence testing and evaluation startup Openlayer Inc. said today it has raised $14.5 million in an early-stage funding.

The money from the Series A round will help it to enhance its platform, which provides enterprises with a host of useful tools to ensure the reliability and efficiency of their AI applications.

Today’s round was led by Race Capital and saw participation from the likes of KPN, Y Combinator, NXTP, Mindset and Quiet Capital. The Spanish telecommunications giant Telefónica S.A. also took part in the round, which comes two years after the startup raised $4.8 million in seed funding.

Openlayer says that in an AI industry where new innovations are being rolled out almost every single week, enterprises require a single, native platform to ensure their products are working as designed, and that is exactly what it provides them with.

According to the startup, AI promises to deliver such explosive growth for enterprises that many are moving extremely rapidly in an effort to gain an advantage over their rivals. But it believes that many companies are deploying AI without proper governance or oversight, and says this can have serious consequences for both the businesses and their end users, should something go wrong.

Openlayer wants to prevent that, and using its comprehensive AI evaluation platform, companies can quickly iterate on new applications, bringing them from prototype to production while reducing the risks associated with AI, such as security vulnerabilities and “hallucinations,” or inaccurate responses.

It was founded by three friends  — Rishab Ramanathan, Vikas Nair and Chief Executive Gabriel Bayomi – who formerly worked together on Apple Inc.’s machine learning efforts. They have built a platform that supports AI developers across the entire application development lifecycle. Designed to cater to both traditional machine learning and generative AI systems, it addresses everything from data quality issues to complex model evaluations and governance, from a single portal.

“When enterprises deploy AI, there’s no room for error, especially in customer-facing applications,” Bayomi said. “A single failure can erode trust, disrupt lives or lead to legal and reputational fallout. That’s why robust evaluation, observability and governance are foundational to responsible AI deployment.”

Using Openlayer, AI teams can not only test and evaluate, but also improve their AI models and datasets in a systematic way, even after they have been shipped out. As an example, it can help developers to verify the integrity of their training and validation datasets, surface meaningful discrepancies between training, evaluation and performance data, and ensure that new models meet their targeted performance benchmarks.

The platform also provides tools for guaranteeing the fairness of model behavior across data subpopulations, and for comparing different versions of models and datasets. Moreover, it helps to explain model behavior by highlighting which parts of the training dataset were used to make a specific prediction.

Openlayer has previously claimed that its platform is used by a number of Fortune 500 companies, and today it named some of them. It said Amdocs Ltd. and Telefonica are using its tools to evaluate their AI models.

Looking forward, Openlayer will use the money from today’s round to expand its enterprise-grade features and scale its go-to-market plans across key industries and global markets.

Featured image: SiliconANGLE/Freepik AI Suite