EN
Agentic AI is driving the next wave of AI innovation. Agentic systems combine multiple AI models that can work together to perform complex tasks beyond the capability of generative AI systems. These systems are capable of autonomously pursuing complex goals and workflows with limited human supervision. To deliver results, these systems need blazing fast interactivity, ability to handle very large inputs, and high-speed model switching.
To accelerate the adoption of agentic AI, SambaNova is announcing the integration of the CrewAI agentic framework with our Cloud API. With this integration, developers can leverage the high performance of the SambaNova cloud to easily orchestrate complex AI agents seamlessly, addressing critical operational challenges and unlocking new possibilities for automation and decision-making.
Together, CrewAI and SambaNova can accelerate building multi-agent AI workflows that deliver many advanced capabilities. Examples of these capabilities include:
- Simplified Workflow Orchestration, including configurations for defining agents and tasks. These can include lead qualification agents that collect personal and company information, validate leads, and score them on criteria like role relevance and market presence. Another example would include educational research where a topic is investigated and high quality, informative content is created, edited, and reviewed for accuracy.
- Enhanced Multi-Agent Collaboration enabled by tools such as SerperDevTool and ScrapeWebsiteTool for data gathering and validation. Structured outputs could be created using the Pydantic model, which would ensure high data quality for tasks like lead scoring and content writing. Examples of structured outputs include a lead data specialist, which gathers key lead information and evaluates its relevance, a cultural fit analyst, which scores leads on alignment with organizational goals, and a planner agent to plan written content from research insights.
- Real-World Scalability and Applications, including automating end-to-end sales pipelines that collect and score leads based on predefined criteria, as well as filtering high-priority leads and engaging them with personalized emails.
A flow visualization feature from CrewAI simplifies debugging and optimization
All of this is made possible by the unmatched speed and efficiency of the SambaNova platform. Powered by SambaNova’s SN40L chip, which uses a dataflow architecture and three-tiered memory design to enable workflows to achieve industry-leading inference performance.
Get started with CrewAI And SambaNova:
- Access the Platform
Sign up for a SambaNova account and get your API key at cloud.sambanova.ai. - Complete the Agentic Sales Pipeline Notebook
Create a multi-agent pipeline with our guided example notebook available in our Integrations folder on Github. Get started quickly with hands-on code demo, preferable to run with Jupyter Notebook. For any more questions about development, reach out in our Community. - Experiment with the Research Assistant
Try out the Educational Research Crew example to see the real-time value of an application using an efficient agentic workflow. Clone the open-source project on GitHub to run it locally and get access to the code for future agentic projects.
- Build with CLI Project Template
Jump in with the helpful project templates accessible through CrewAI’s CLI. For example, you can run the following lines whether you want a crew or a flow:‘crewai create crew my_new_crew’And then select SambaNova as a model provider and the model you want to use.
‘crewai create flow my_new_flow’
Get started today on SambaNova Cloud and see how CrewAI with SambaNova can transform your AI workflows. To get started with CrewAI, refer to their documentation for templates and quick start guides.
