Carbon Perplexity's integration will make it easier for enterprises to connect their data with AI analysis

2024 has been a banner year for Anxiety. AI research startup, founded by former DeepMind and OpenAI researcher Aravind Srinivas, raised hundreds of millions of dollars – its latest funding round reported valuing the company at $9 billion – and introduced several unique features, including Pages, Locationsand innovative shopping experiences.

These developments have cemented Perplexity's reputation as an “AI-first” knowledge search engine, standing apart from traditional search giants such as Google and Bing, which incorporate AI capabilities into their existing engines.

However, the journey is far from over.

Facing more intense competition, Perplexity expands its range with a new extension to its portfolio: Carbon. The company just has received this startup, for an undisclosed amount, to address the “data gap” that enterprises encounter with AI research and streamline the knowledge discovery process in their workflows.

Carbon has developed a comprehensive retrieval framework that streamlines the process of linking external data sources to LLMs. Users can tap the Carbon Universal API or SDKs to synchronize their data sources and retrieve the data for use by LLMs. It offers native integration with over 20 data connectors and supports over 20 file formats, including text, audio and video files.

The expanding scope of AI research

From individuals to business users, almost everyone today uses AI analysis as part of their workflows. The idea of ​​the technology is very simple – you don't have to wade through a bunch of links and content to find relevant insights and information. Instead, the information comes to you as a direct response to your question.

Uncertainty has found success with this approach, using a range of large language models to retrieve information from the web and simplify how users work. It even allows teams to pull information from their personal or business files such as PDFs and Word documents.

But, here's the thing. The web is home to public information, and uploading internal files – PDFs, conversations, images – alone is not feasible for business users dealing with large amounts of proprietary data. This affects the quality of answers, keeping them general and without important contexts that belong to an organization.

Highlighting this “data gap,” Sanjeev Mohan, Gartner Research's former vice president for data and analytics, told VentureBeat that ETL is one of the biggest AI trends for 2025. unstructured data. It allows teams to extract and transform data from distributed internal sources, ultimately empowering their LLMs to generate highly relevant and accurate responses.

Now, this is exactly what Perplexity plans to do with Carbon's comprehensive, streamlined recruitment framework. Complexity integrates Carbon's retrieval engine and connectors into its technology stack, giving users of the search platform a direct way to deploy their diverse data sources. in, from Google Docs and Notion to Hubspot and Slack.

This, the company says, will expand the pool of knowledge that powers the AI ​​search engine, making its answers more complete, relevant and personalized to users.

What can consumers expect from carbon-powered Perplexity?

While Perplexity has just received Carbon and the integration has yet to be executed, it is very easy to imagine how the additional data connections will improve the workflows of enterprise teams using the AI ​​search engine.

For example, if one needs to move the launch date and needs to work out the latest date and direction set by their team, Perplexity would be able to check the all data in Google Docs, Notion, and Slack – and use it. correlations – to find the information that answers the question.

In fact, there would be no more worries about stitching together context from the web, individual apps, and messages. The platform does everything by itself to provide the answer.

“The unique advantage of this setup is that our technology can get the answer without forcing you to identify the document/database where that information is stored,” said Sara Platnick, who is head of communications at Perplexity, to VentureBeat.

Another example, she said, could be producing customer meeting scenes. Your concern would be able to get the details and focus of the conversation from connected CRMs in no time.

In particular, by benefiting from carbon improved generation-detection (RAG), Perplexity makes enterprise analytics more accessible, saving companies the hassle of building their own RAG pipelines from scratch.

“By discovering and interpreting proprietary data with Perplexity and Carbon, companies can address a range of multi-faceted gen AI use cases. We find that the primary adopters are focused primarily on customer service, document processing, image processing and recommendation engines,” Kevin Petrie, VP of research at BARC US, told VentureBeat.

Performance will be a priority

Getting Carbon is just the beginning. The real key is performance, or how quickly and safely the startup's technology will be integrated. After all, we are talking about proprietary data from some of the most critical sources of knowledge that enterprises maintain.

“Companies are rightly wary of revealing their intellectual property to the public. Perplexity and Carbon must therefore provide regulatory controls that ensure companies can keep their data within their own firewalls. They are not interested in sharing secrets or training a public model to simulate their intellectual property,” said Petrie.

On the Perplexity side, Platnick noted that “all information from internal and private sources on the engine is encrypted, as is all data transferred and stored in link Carbon data.” She also pointed out that the company has additional safeguards to ensure that private documents remain private and are not accessible to unauthorized users.

As of now, there is no specific timeline for the integration of Carbon with Concern. However, the startup will stop working on its managed API on March 31, 2025. Customers already using the API have been contacted to get on board, with the Carbon team assisting them in the transition.