AI

Scaled Cognition nabs $100M to automate high stakes customer interactions

Artificial intelligence startup Scaled Cognition Inc. today announced that it has raised $100 million in funding.

Khosla Ventures led the Series A deal with participation from Genesys Telecommunications Laboratories Inc., a provider of customer service software. The Wall Journal reported that Scaled Cognition is now worth $850 million.

Founded in 2022, Scaled Cognition sells an AI platform that enterprises can use to process customer service requests. The company says that the software is optimized to complete user-specified actions correctly on the first attempt. That makes it suitable for sensitive tasks such as making purchases on a customer’s behalf. 

Scaled Cognition’s platform runs on a custom AI model called APT. It doesn’t generate new data in response to user requests but rather retrieves existing records, an approach that the company says can reduce the risk of hallucinations. For added measure, APT double checks that it completed a task before displaying a confirmation message.

Customers can further reduce the risk of output errors using a simulation tool that Scaled Cognition ships with the model. APT retrieves the data it requires for prompt responses from companies’ backend systems. According to Scaled Cognition, its simulation tool creates a replica of those systems’ application programming interfaces and tests how reliably APT interacts with them.

A browser-based wizard enables workers to create AI agents for specific support use cases. Tech-savvy companies can use a software development kit to extend their agents, add safety guardrails and run reliability tests.

Scaled Cognition also provides a third AI agent development tool. AgentTwin, as it’s known, can automatically create an agent based on historical customer interactions and data about a company’s most performant existing agents. According to Scaled Cognition, version control and testing tools enable software teams to refine their agents over time.

The company ships its platform with a cybersecurity module that blocks malicious prompts. For added measure, the software collects telemetry about AI agents’ customer interactions. Enterprises can use the data to identify and fix agent reliability issues.

“Reliability is engineered into the architecture of our models, not bolted on after the fact,” said Scaled Cognition co-founder and Chief Technology Officer Dan Klein. “The biggest reliability challenge isn’t the mistakes that look wrong; it’s the ones that look completely correct. If you want AI to take real actions on behalf of customers, that’s the problem you have to solve.”

The company will use its newly closed funding round to accelerate its research and commercialization efforts. In the longer term, Scaled Cognition plans to extend its platform’s focus beyond customer service to other use cases such as finance.

Image: Scaled Cognition