Zapier CEO: Most AI agent talk is hype

This article summarizes an episode of SaaStr AI’s video series featuring CEO of Zapier Wade Foster.

CEO Zapier Wade Foster / Photo credit: Wade Foster’s Twitter

The hype around autonomous AI is misleading. Most companies aren’t deploying fully independent AI systems, known as agentic systems, across their operations. Wade Foster, co-founder and CEO of the automation platform Zapier, explained this on the SaaStr AI Podcast, an event focused on software-as-a-service companies.

His insights point to a shift in how businesses operate. Automation is moving from simple integrations to a hybrid model. This approach combines predictable, rule-based (deterministic) AI with more flexible, goal-oriented (agentic) AI, with humans stepping in for final, high-stakes decisions to ensure accuracy.

Automation has moved beyond linear tasks

Wade Foster described how automation has evolved. The technology started with simple trigger-action connections linking two applications. It then advanced to multi-step workflows that could chain several actions together, paving the way for AI decision-making in core processes.

The first phase was simple integration
Foster explains that his company’s platform started as a tool for simple integrations, like connecting a website form to a customer relationship management (CRM) system.

Foster states, “the Zapier that people are probably knowing familiar with was a simple integration tool. It’s like trigger action. You get a lead from a form on your website, add it to CRM.”

The second phase embraced multi-step workflows
He then describes the move to more complex workflows that could tap into an enrichment API, a service that adds more data to existing records.

Foster continues, “phase two of Zapier was then embracing full-on workflow automation. So instead of single trigger, single action, you got trigger multiple actions. So you could say, ‘Hey, a lead comes in on my website, let’s go hit an enrichment API and then let’s add it to our CM.'”

The current phase embeds AI directly into processes
Foster explains, “now with AI, you’re able to embed AI as part of those workflows end to end. and so you can say, ‘hey, new lead comes into my site. Hit an enrichment API, but also go do deep research on this customer.'”

Agentic systems unlock non-linear problem solving

This evolution leads to agentic systems that operate differently from structured workflows. Unlike predefined automations, these agents can interpret goals and determine their own course of action, which is a new way to execute complex, non-linear tasks.

Agentic workflows are defined by choice, not pre-set paths
Foster clarifies, “[Zapier agents are] not even a workflow at all. It’s an agent that really has the choice to figure out how do you want to go do this. So you can say, ‘hey agent, reply to my emails.'”

Agents become more effective with detailed instructions
Foster notes, “now you probably want to give [the agent] more instructions than just that. But the more instructions you give it, the more the agent gets smarter about how to reply to email.”

Fully autonomous agents are overhyped

While agentic systems offer powerful new capabilities, Foster cautions against market hype. He argues that the excitement around fully autonomous agents is largely speculative. The most effective operators are not using pure agentic systems for every business process.

The most effective automation blends predictability and flexibility

AI handles the prep work for human decisions