Join top executives in San Francisco July 11-12 to hear how leaders are integrating and optimizing AI investments for success. Learn more
Galileo, a San Francisco-based AI startup, today announced the launch of Galileo LLM Studio, a platform for diagnosing and solving problems with large language patterns. The platform aims to help companies deploy natural language processing models in manufacturing faster by detecting “model hallucinations” or incorrect predictions and improving model accuracy.
In an exclusive interview with VentureBeat, Galileo co-founder Yash Sheth explained the vision behind the LLM Studio: “We truly believe that Generative AI is poised to change the world. Businesses, governments and individuals can now finally interact with AI in ways that were not possible with predictive machine learning.”
The platform comes as demand for natural language processing has skyrocketed, with companies eager to use templates for applications such as chatbots, intelligent search and automatic text generation. However, creating and implementing these complex models remains a challenge. According to Sheth, data scientists spend a large portion of their time “data cleaning,” fixing problems in datasets to improve model accuracy.
“Even though we had the best talent, the best team, the best infrastructure, it took months to get a model into production,” said Sheth, reflecting on nearly a decade of working on machine learning at Google. “When we started looking out, this was the status quo in the AI industry.”
Accelerate the pace of adoption
Galileo’s platform aims to automate much of the work normally involved in cleaning datasets. Galileo Prompt Studio detects “model hallucinations” or incorrect predictions, enabling data scientists to fix errors faster. The platform also allows data scientists to compare multiple requests to find the optimal input, and estimates the cost of calls to external AI services like OpenAI to help manage budgets.
With generative models becoming increasingly commoditized, Sheth believes the key to unlocking their potential lies in understanding the impact of data and adapting these models. “It takes a long time to really adapt these models and make them work. Anything we can do to accelerate will only accelerate the adoption of AI around the world,” she said.
The startup also hopes to expand beyond natural language processing to other AI domains such as computer vision. “Our algorithms span all data formats, because ultimately we embed them in neural networks, and the neural networks’ representation of the data is just a vector of floats,” said Sheth.
With $18 million in funding from investors including Battery Ventures, Galileo is poised to capitalize on the growing demand for handy AI tools. However, the company faces stiff competition from tech giants like Google, Microsoft, and AWS, which also offer platforms for building and managing AI models. Galileo hopes his focus on diagnosing and correcting model errors will differentiate them.
“Being data-centric and having a diagnostic view of the key model throughout the machine learning lifecycle is absolutely critical to AI adoption,” said Sheth.
VentureBeat’s mission it is to be a digital city square for technical decision makers to gain insights into transformative business technology and transactions. Discover our Briefings.
