Highlights:

  • Dappier runs a marketplace where publishers can offer their content to developers for integration into AI projects.
  • The company states that its marketplace supports AI models employing RAG, which stands for retrieval-augmented generation.

A startup that assists publishers in licensing their content to artificial intelligence developers, Dappier Inc., has secured USD 2 million in seed funding.

Silverton Partners, an early-stage venture capital firm, led the investment.

Dappier, headquartered in Austin, runs a marketplace where publishers can offer their content to developers for AI projects. The business charges clients per query under a pay-as-you-go pricing structure. Pricing for queries is determined by the publisher of the dataset being accessed.

In addition to offering monetization tools, Dappier also simplifies the technical processes required to make data accessible for AI projects.

Before a collection of records can be fed into neural networks, it needs to be structured into a format that is easier to process. Dappier aims to automate this process by offering a series of prepackaged data models. These models are files that specify how the individual records within a dataset should be structured.

Publishers can easily upload their content to Dappier’s marketplace using a no-code interface. The platform pulls content from documents, databases, RSS feeds, and Airtable, a popular cloud-based collaboration platform. Dappier automatically arranges these files for processing based on its pre-installed data models after they are uploaded.

The company states that its marketplace supports AI models utilizing RAG, which is short for retrieval-augmented generation. RAG is a machine learning technique enabling neural networks to retrieve data from external systems. This approach allows developers to incorporate new information into their AI models without needing to retrain them, thereby lowering costs.

Dappier claims that its platform can streamline publishers’ internal AI projects. Website operators who monetize their data through the platform utilize it to develop AI-driven search and content recommendation functionalities for their audience.

The money raised in Dappier’s seed round will be used for product development. The business intends to give publishers additional options for uploading their content to its platform. It will also help to increase the quantity of data that AI developers have access to.

Co-founder and Chief Executive Officer Dan Goikhman said, “Dappier Marketplace allows publishers to easily make their data accessible to AI agents, and then license directly – ensuring fair compensation for the content that fuels improved AI experiences.”

Content licensing has become increasingly significant for AI providers. Last year, OpenAI secured agreements with Politico and Insider’s parent company, Axel Springer SE, to provide news articles for its large language models. Subsequently, the developer of ChatGPT has also disclosed similar licensing arrangements with Vox Media Inc., The Atlantic, and various other publishers.

AI developers are actively seeking additional training data for their language models, not limited to news content. OpenAI and Google LLC, for instance, are each reportedly paying Reddit Inc. around USD 60 million annually for access to its social network data. These businesses have also signed separate license agreements to use Stack Overflow’s library of coding tips.