Now, Airbyte's data platform can build API connections by reading the documentation

If your startup has any involvement with data pipelines, you’re likely trying to take advantage of the current trend: Enterprises are exploring how to harness data to power generative AI products, which requires robust data infrastructure. Airbyte, founded in 2020, initially focused on creating a low-code/no-code open-source data integration platform. Since then, the company has raised a total of $181.2 million, including a substantial $150 million Series B round in the somewhat atypical market of late 2021.

Now, after four years, Airbyte is introducing Airbyte 1.0, emphasizing AI as a core component, both to enhance its own tools and to support users in developing AI-based services.

The company is cleverly using AI to advance its low-code/no-code approach: Its model can examine API documentation and automatically generate a connector from it. Users simply direct the model to the documentation, and it handles the rest—though its effectiveness in real-world scenarios remains to be seen.

According to Airbyte co-founder and CEO Michel Tricot, large language models (LLMs) are revolutionizing how enterprises use data by making unstructured data significantly more accessible and useful.

“Structured data only represents a fraction of the potential of data,” Tricot explained. “With LLMs, we can now effectively utilize unstructured data that was previously untapped. We’ve seen a high demand for managing multi-modal data. Our recent efforts have focused on supporting intelligent, context-aware pipelines, optimizing frameworks like RAG, and automating pipeline creation based on user workflows. These advancements are essential for enabling advanced use cases and improving LLM performance.”

Airbyte’s enhanced ability to manage unstructured data means that users can now leverage their existing pipelines for this purpose without the need for additional tools.

In non-AI developments, Airbyte’s connector now also supports GraphQL, allowing users to access more datasets without having to create custom pipelines.

With this release, Airbyte is also making its self-managed enterprise service widely available. Similar to many open-source companies, the enterprise version, available on AWS and GCP marketplaces, includes features such as single sign-on (SSO), role-based access control (RBAC), and unique Airbyte features like sensitive data masking and advanced observability.

Airbyte reports having 7,000 enterprise customers and over 170,000 deployments to date. Its clients include well-known names like Calendly, Coupa, Perplexity AI, and Siemens.

“Every company is a data company, driving decisions and forming the backbone of AI initiatives,” Tricot said. “Only Airbyte, with our open-source strategy offering hundreds of connectors, allows enterprises to use any data they need. As AI continues to transform industries, we’re providing the technology and ecosystem necessary for organizations to build the data infrastructure required for AI-driven innovation.”