Agentblazer: Salesforce announces agentic AI certification, learning path

Hot on the heels of the release of Agentforce 2dx for developing, testing, and deploying AI agents, Salesforce on Thursday introduced Agentblazer Status to its Trailhead online learning platform. Agentblazer Status recognizes three status levels in the development of Agentforce skills and culminates in the new Agentforce Specialist certification.

“Our goal is to enable 1 million agent pioneers we call Agentblazers,” Adam Evans, EVP and GM of Salesforce AI, said at a press conference on Monday. “We’re announcing new skills, new certifications, and a slew of new features to kickstart this next round of innovation.”

Evans explained that agentic AI requires businesses to think differently about software development.

“Traditional software building is rigid,” he said. “It’s clicks, it’s code, it’s logic.”

But AI agents can observe, reason, and act autonomously.

“The way we build with agents is fundamentally different from the way we used to build software,” added Alice Steinglass, EVP and GM of Salesforce Platform.

Before AI agents, Steinglass said, building a hotel check-in, for example, would require writing every “if-then” statement, accounting for every possibility and hardcoding it.

“Today, when I think about interpreting that into instructions for an agent, my first gut instinct is to write down every single thing that could possibly happen — what if they’re a vegetarian or need a different room or want a view of the lake?  All of that needed to be hardcoded,” she said. “But now, with agents, I’m doing it differently.”

Instead of accounting for every possibility, she said she would break it down into smaller chunks, describe the action she would want the agent to take in the scenario, and then allow the agent to dynamically compose those individual pieces together at runtime.

“I can bring in more flexibility and understanding so my agent can interpret those instructions,” she said. “I could have a policy document that explains them instead of hardcoding them with if-then statements.”

To help its customers develop the skills they need to build, deploy, and manage agentic AI, Salesforce has created the new Agentblazer AI learning roadmap through Trailblazer. The company said Agentblazer Status recognizes Agentforce skills and provides a clear learning path for building autonomous agents and showcasing expertise. The statuses are:

• Agentblazer Champion (level 1): Agentblazers with this status learn AI and Agentforce fundamentals, as well as building understanding of Data Cloud and governance. This level includes helping Agentblazers build their first AI agent.
• Agentblazer Innovator (level 2): This status builds on the champion level, focusing on identifying business use cases and building custom agents with Agent Builder. It teaches developers how to configure Agentforce for Service and Sales.
• Agentblazer Legend (level 3): The top-level status requires a deep understanding of advanced concepts. Agentblazer Legends need to design complex strategies for managing and customizing the Agentforce lifecycle within their organizations.

Each of the three levels is associated with dedicated Trails, including modules and hands-on challenges involving AI and Agentforce. Learners earn badges for their Trailblazer Profile by completing the modules.

Salesforce said attaining the Agentblazer Legend status prepares developers to earn its Agentforce Specialist certification.

As part of its push to get to 1 million Agentblazers and its AI for All initiative, Salesforce said that a candidate’s first attempt at the Agentforce Specialist exam will be free for all through the end of 2025.

[ See also: Salesforce certification guide: Roles, paths, exams, cost, training, requirements ]