What is Scaled Cognition and what does it aim to solve in AI?

Scaled Cognition is an AI model lab based in Mountain View, California, that aims to solve the challenge of reliability in AI.

Who led the Series A funding round for Scaled Cognition and how much was raised?

Khosla Ventures led the Series A funding round for Scaled Cognition, raising $100 million.

How does Scaled Cognition's approach differ from its rivals?

Scaled Cognition's approach differs from its rivals as it prioritizes reliability in the architecture of its models, rather than wrapping a safety layer around an existing frontier model.

What is Scaled Cognition planning to do with the funding they raised?

Scaled Cognition plans to use the funding to expand its research team and accelerate enterprise deployments.

Scaled Cognition Raises $100M for AI Model Designed to Eliminate Hallucinations

Q: What industries is Scaled Cognition currently working with?

Scaled Cognition is already in production with Fortune 500 enterprises across financial services, healthcare, telecom, and insurance.

AI model lab Scaled Cognition secures Series A funding led by Khosla Ventures, focusing on improving reliability in AI systems. The company's flagship model aims to reduce errors in critical enterprise workflows.

Scaled Cognition, an AI model lab based in Mountain View, California, has raised $100 million in a Series A funding round led by Khosla Ventures. The company aims to solve one of the biggest challenges in AI today: reliability. While AI systems have become remarkably capable, they remain prone to mistakes and hallucinations that limit their use in critical enterprise workflows.

The funding will be used to expand Scaled Cognition's research team and accelerate enterprise deployments. The company is already in production with Fortune 500 enterprises across financial services, healthcare, telecom, and insurance - where a wrong answer has real consequences.

Background and Context

Scaled Cognition was co-founded by CTO Dan Klein, a UC Berkeley professor of AI and leading natural language processing researcher, and CEO Dan Roth, who together previously built and sold one of the first agentic AI companies to Microsoft. The company's core claim is that reliability cannot be bolted on after the fact; instead, it must be designed into the architecture of the model.

This approach sets Scaled Cognition apart from rivals who wrap a safety layer around an existing frontier model. The company has spent years rebuilding its architecture to prioritize reliability, and its flagship model, APT (Agentic Pretrained Transformer), is designed to match the conversational quality of leading models while eliminating hallucinations and sticking to policy.

Why it Matters to the Industry

The industry's reliance on AI has grown exponentially in recent years. However, as AI systems become more widespread, their limitations are becoming increasingly apparent. The inability to guarantee reliable performance is a major hurdle for enterprises looking to integrate AI into critical workflows.

Scaled Cognition's solution addresses this issue directly. By prioritizing reliability and designing it into the architecture of its models, the company aims to provide a level of trustworthiness that is currently lacking in the industry. This has significant implications for industries where errors can have real consequences, such as financial services, healthcare, and telecom.

What Comes Next

The funding will be used to expand Scaled Cognition's research team and accelerate enterprise deployments. The company is already working with several major enterprises, including Genesys, a cloud contact-centre giant serving more than 8,000 organisations in over 100 countries.

Scaled Cognition aims to automate more than one billion customer support interactions within the next twelve months. The company's models are designed to resolve most issues, saving "hundreds of millions" in operational costs, according to CEO Dan Roth.

Key Facts

Scaled Cognition has raised $100 million in a Series A funding round led by Khosla Ventures.
The company aims to solve the challenge of reliability in AI models.
Scaled Cognition's flagship model, APT (Agentic Pretrained Transformer), is designed to match the conversational quality of leading models while eliminating hallucinations and sticking to policy.
The company is already working with several major enterprises, including Genesys.
Scaled Cognition aims to automate more than one billion customer support interactions within the next twelve months.