Perplexity has launched Teammates, a new suite of AI agents designed specifically for workplace collaboration and project management. These always-on digital workers can assume multiple roles within a single workspace while accessing existing projects, skills, and memory from the Perplexity tools users currently employ. The system integrates with Perplexity Computer, an existing engine that deploys agents directly on a user’s device to execute tasks. Perplexity states that Teammates build upon this foundation by enabling collaboration on ongoing, large-scale work rather than isolated actions. Users can assign distinct jobs to individual Teammates, allowing them to perform tasks simultaneously in a manner similar to OpenAI’s Dots or Anthropic’s Claude Duet. The service aims to provide a structured environment where these agents coordinate efforts across complex workflows without requiring separate manual intervention for every step.
The significance lies in the shift from single-purpose bots to coordinated teams that maintain context over time. By leveraging existing user data and tool access, Perplexity attempts to reduce the friction of setting up new AI workflows for every project. This approach targets professionals who need sustained assistance on long-term goals rather than quick, one-off queries. The company is positioning itself against competitors who have recently released similar agent frameworks by emphasising deep integration with its own search and analysis infrastructure. Success will depend on whether these agents can reliably manage complex, multi-step processes without hallucinating instructions or losing track of previous context. The rollout represents another step in the broader trend of moving from chat-based interactions to autonomous execution environments.
- Teammates access existing projects and skills from Perplexity tools
- Agents can perform multiple simultaneous jobs within a workspace
- Integration relies on Perplexity Computer’s on-device deployment capabilities
