ADEPT

AI Focus Areas

• Action-taking AI agents
• UI and API automation via language models
• Enterprise workflow orchestration
• Code and data tool integration
• Human-computer collaboration research

Key Products

• Adept action model (internal)
• Agentic assistants that operate SaaS tools
• Enterprise automation workflows
• APIs for integrating action-taking AI

Market Position

Adept focuses on a specific interpretation of AGI: systems that can use existing software like a human operator, rather than replacing it. This agent-centric view differentiates it from pure LLM providers. With backing from General Catalyst, Spark Capital, Greylock, Microsoft, Nvidia and others, Adept sits at the intersection of infrastructure and application, integrating deeply into enterprise tools. It competes with copilots from Microsoft, Google and startups building agent platforms, but its research agenda and early start on training models to act in complex interfaces give it an edge. The challenge is turning impressive demos into robust, safe and monetizable enterprise products across many heterogeneous systems.

AGI Relevance

Using tools and software interfaces is a core capability on the path to AGI. Adept’s models observe screens, interpret UI elements and trigger actions in sequence, effectively learning to ‘drive’ existing digital environments. This moves AI beyond text prediction toward embodied competence in the digital world. Training such agents raises key AGI questions: how to represent goals and constraints, how to avoid harmful or irreversible actions, and how to learn from weak supervision in messy interfaces. Adept’s research and datasets on UI interactions will inform how future AGI systems interact with software, especially in enterprise contexts where fully replacing tools is unrealistic.

Investment Highlights

Raised $65M+ before 2023; in March 2023 announced a $350M Series B co-led by General Catalyst and Spark Capital, with participation from Addition, Greylock, Atlassian Ventures, Microsoft, Nvidia and others, at a valuation reported to be at least $1B.

Tags

• Action Models
• Automation
• Agent