Helping Enterprises Build and Improve AI at Speed and Scale
Braintrust Data, a startup based in California, has successfully raised $5.1 million in a seed round of funding. The funding was led by Greylock Partners, a well-known venture capital firm. Braintrust Data, founded just over two months ago by Ankur Goyal, aims to address the challenge of AI evaluation in enterprises by providing a dedicated tool that allows teams to assess and enhance their AI models before they reach the production stage.
Despite being a relatively new company, Braintrust Data has already attracted numerous customers and investments from prominent figures in the industry including Elad Gil, Clem Delangue, Greg Brockman, Jack Altman, Howie Liu, Guillermo Rauch, Bryan Helmig, Simon Last, and Vipul Ved Prakash. The company plans to utilize the funding to expand its team and further develop its platform, allowing developers to work more efficiently and stay at the forefront of AI.
The Challenge of AI Evaluation in the Dev Loop
AI plays a crucial role in modern business applications. However, ensuring that these applications perform at their best can be a complex task. Even a small code change aimed at improving an application can inadvertently break the entire workflow. This reactive approach to bug fixing can negatively impact the customer experience.
“Our product allows you to easily (in under an hour) instrument your code to define evaluations, capture user feedback, log LLM calls, etc. Every time you make a change, you can re-run evaluations and instantly get a dashboard that tells you how much you improved or regressed things, and debug individual cases (before moving to final deployment). You can also log examples from staging/production and run evaluations against them to find new edge cases users are hitting,” Ankur Goyal, CEO of Braintrust Data, explained.
To address this challenge, developer teams focus on the practice of evaluation in the dev loop. They analyze context-specific data and metrics, and experiment with various models, prompts, and fine-tuning techniques to achieve the desired results. However, this process can be time-consuming and delay the launch of new features.
Ankur Goyal, drawing from his experience at his previous ventures Impira and Figma, recognized the need for a solution that enables faster evaluations. He launched Braintrust Data in August 2023, and it has already gained traction with “hundreds” of enterprises and startups as customers, including Airtable, Zapier, Coda, and Instacart. According to Goyal, these customers have experienced a significant improvement in the accuracy of their AI offerings with the help of Braintrust, resulting in faster ship cycles, increased engagement, and better team collaboration.
“Our product can run inside of your own cloud environment, which is critical for enterprise security, especially in AI which is rampant with PII and proprietary info. This has enabled our enterprise customers to use Braintrust for their most mission-critical workloads,” Goyal added.
Expanding Capabilities and Differentiation
Besides evaluation capabilities, Braintrust has started offering additional features to help AI teams iterate and ship faster. These include a prompt playground to compare multiple prompts, benchmarks, dataset management, and an AI proxy providing access to popular AI models.
In an increasingly competitive market, Braintrust differentiates itself by offering insights before the model reaches the production stage.
“There is no doubt this is an exciting space with other companies trying to add value. Most products out there are focused on observability, which allows you to see what’s happening in production. Unfortunately, if you only have observability, you have to ship things to your users to find out whether they work. We’ve found that engineering teams who implement great evaluations move significantly faster – up to 10 times faster – than those who are just watching what happens in production and trying to fix them ad-hoc,” Goyal emphasized.
With the latest funding round, Braintrust has raised a total of $8.3 million. The company plans to utilize these funds to hire additional talent and continue developing its platform to become the market-leading solution for evaluations and support advanced AI tooling.