Galileo Technologies Inc., a provider of enterprise AI observability and assessment platforms, today announced that it has raised $45 million in new funding.

The Series B round was led by Scale Venture Partners to help companies build trustworthy AI platforms. Databricks Ventures, Premji Invest, Amex Ventures, Citi Ventures, ServiceNow and SentinelOne participated alongside existing investors Battery Ventures, Walden Capital and The Factory. Additionally, Andy Vitus, partner at Scale Venture, will join Galileo’s board of directors.

Large language models and applications built on them depend on precise answers. Today’s LLMs are becoming larger, more complex and more powerful than ever before. and that means that going from demo to production ready is not easy while maintaining this accuracy. Galileo has created an all-in-one assessment and observability stack that provides metrics and protection against common issues.

“We launched Galileo three years ago to solve the AI ​​measurement problem, particularly focusing on language models,” said co-founder and CEO Vikram Chatterji. “Using humans or LLMs to evaluate model responses is expensive, slow, and does not scale. Yet today, these are the de facto techniques adopted by AI teams.

Chatterji says Galileo differentiates itself with a research-based approach and a user experience specifically designed to help enterprise users. manage Assessment and confidence in LLM and LLM applications at the scale they need. According to Gartner Inc.the progress adoption rate means more than 80% of businesses will include generative AI by 2026.

Company says AI has a ‘measurement problem’ Or obtaining appropriate metrics on AI performance is difficult and requires specialized tools. To manage this, the company developed what is it called the assessment intelligence platform that integrates assessments into the AI ​​development pipeline so teams can rigorously test LLMs, train and refine, debug during experimentation, and secure after deployment.

This became the basis of the company’s flagship product, the Luna Evaluation Suitea collection of high-performance benchmarking models designed to be accurate, fast, and cost-effective. Out of the box, they can evaluate the performance of LLMs such as OpenAI’s GPT-4o and Google LLC’s Gemini Pro, without requiring ground truth. They can perform a large number of tests and gather various information without having to prepare tedious test sets.

“Traditional assessment methods, such as human assessments or using LLMs as judges, are not suitable for enterprise use cases,” said Jim Nottingham, senior vice president and president of the Advanced Computing Solutions Division at HP Inc. “Galileo’s rapid innovation and commitment to overcoming the biggest hurdles in assessment – ​​from accuracy to bias – provide a comprehensive view of generational AI application performance, which is why they have become an essential part of our Z by HP AI Studio.

Given the growing adoption of AI language models, Galileo said that since the start of 2024, the company has brought on six Fortune 50 companies as customers, including Comcast Corp. and Twilio Inc.

Chatterji said the new funding go deeper Galileo’s development and engineering efforts to meet market demands for more accurate and reliable AI systems and to expand its product portfolio.

Image: Galileo