Collect and summarize async standup updates
Lessons from Zapier’s experience building hundreds of internal AI agents to reduce engineering friction. Task: Lessons from Zapier’s experience building hundreds of internal AI agents to reduce engineering friction. Justin Reock Deputy CTO This post was originally published in Engineering Enablement, DX’s newsletter dedicated to sharing research and perspectives on developer productivity. Subscribe to be notified when we publish new issues. Across engineering organizations, AI is often framed as a cost-cutting tool: automate work, reduce headcount, and run leaner teams. But as more companies move beyond pilots, a different pattern is emerging. When AI is applied to the right problems, it doesn’t shrink engineering teams. It increases the return on every engineer you hire. Zapier is a clear example of this shift. I recently had the chance to interview Andrew Kordampalos, who leads Zapier’s AI Agents transformation. Zapier has built an internal ecosystem of lightweight AI agents that automate thousands of operational and administrative tasks across engineering, product, and internal operations. Their experience points to a simple conclusion: as AI increases the value of each engineer, it becomes more rational to hire more engineers, not fewer. This article looks at how Zapier reached that conclusion and how they built a culture that treats AI as a force multiplier for hiring and value creation, rather than as a headcount-reduction strategy. Using AI to remove friction around engineering work Zapier’s starting point was not “How do we replace developers?” but “Where does work slow developers down?” When Kordampalos joined, his team had grown out of a startup (acquired by Zapier) that originally built real-time meeting transcript agents. Their team quickly turned towards a bigger question: Would it be more effective to automate the choreography around engineering rather than the core engineering itself? With this line of thinking, they formed a hypothesis that engineers would become far more productive if the friction around them dissolved through agentic automation. The goal was not to change what engineers do, but to change how much of their time is spent on work only they can do. Zapier’s AI Agents team introduced a network of lightweight internal agents that automate the coordination and overhead surrounding development work: Agents that collect async standup updates and summarize them, allowing teams to replace five daily standups with two weekly sessions. Agents that run onboarding steps, including generating email signatures and triggering access to tools, reduce onboarding time to roughly two weeks—significantly less than the industry standard 30–90 days. Agents that manage Slack workflows, approvals, and operational routines, cutting down on context switching and interruptions. The underlying mindset is simple: “When you find you’re doing the same thing again and again, you build the automation.” Zapier already had a culture of “build the robot.” AI agents gave that culture a larger surface area. How AI agents changed Zapier’s hiring math As Zapier deployed more internal agents, they saw a compounding effect. New hires reached productivity faster. Existing employees increased their throughput without extending their working hours. The organization started to experience AI not as a way to substitute for engineers, but as infrastructure that made every engineer more effective. This is where the economics shift. If AI removes 10–15% of the administrative and coordination work from an engineer’s week, their effective output goes up. The cost of hiring remains the same, but the value per hire increases. In that scenario, reducing headcount undercuts your own leverage. Zapier leaned into this logic. As Kordampalos puts it, “We are doubling down and hiring even more people because we want to boost our productivity by using AI.” AI didn’t replace engineers; it replaced the parts of engineering work that prevent engineers from doing their best engineering. The move from summarizing meetings to a full agent ecosystem; tools: Zapier Agents. Human role: The original post does not say how a person stayed involved.; limitations: The source does not detail specific integration costs, error handling protocols, or the exact technical stack beyond 'Zapier Agents'. It assumes a culture of 'build the robot' and relies on Slack as the coordination layer. Human approval is implied for governance but not explicitly defined as a mandatory step for every message..
Tools
Published on Agent Work Atlas:
Overview
Lessons from Zapier’s experience building hundreds of internal AI agents to reduce engineering friction.
Workflow
Lessons from Zapier’s experience building hundreds of internal AI agents to reduce engineering friction. Justin Reock Deputy CTO This post was originally published in Engineering Enablement, DX’s newsletter dedicated to sharing research and perspectives on developer productivity. Subscribe to be notified when we publish new issues. Across engineering organizations, AI is often framed as a cost-cutting tool: automate work, reduce headcount, and run leaner teams. But as more companies move beyond pilots, a different pattern is emerging. When AI is applied to the right problems, it doesn’t shrink engineering teams. It increases the return on every engineer you hire. Zapier is a clear example of this shift. I recently had the chance to interview Andrew Kordampalos, who leads Zapier’s AI Agents transformation. Zapier has built an internal ecosystem of lightweight AI agents that automate thousands of operational and administrative tasks across engineering, product, and internal operations. Their experience points to a simple conclusion: as AI increases the value of each engineer, it becomes more rational to hire more engineers, not fewer. This article looks at how Zapier reached that conclusion and how they built a culture that treats AI as a force multiplier for hiring and value creation, rather than as a headcount-reduction strategy. Using AI to remove friction around engineering work Zapier’s starting point was not “How do we replace developers?” but “Where does work slow developers down?” When Kordampalos joined, his team had grown out of a startup (acquired by Zapier) that originally built real-time meeting transcript agents. Their team quickly turned towards a bigger question: Would it be more effective to automate the choreography around engineering rather than the core engineering itself? With this line of thinking, they formed a hypothesis that engineers would become far more productive if the friction around them dissolved through agentic automation. The goal was not to change what engineers do, but to change how much of their time is spent on work only they can do. Zapier’s AI Agents team introduced a network of lightweight internal agents that automate the coordination and overhead surrounding development work: Agents that collect async standup updates and summarize them, allowing teams to replace five daily standups with two weekly sessions. Agents that run onboarding steps, including generating email signatures and triggering access to tools, reduce onboarding time to roughly two weeks—significantly less than the industry standard 30–90 days. Agents that manage Slack workflows, approvals, and operational routines, cutting down on context switching and interruptions. The underlying mindset is simple: “When you find you’re doing the same thing again and again, you build the automation.” Zapier already had a culture of “build the robot.” AI agents gave that culture a larger surface area. How AI agents changed Zapier’s hiring math As Zapier deployed more internal agents, they saw a compounding effect. New hires reached productivity faster. Existing employees increased their throughput without extending their working hours. The organization started to experience AI not as a way to substitute for engineers, but as infrastructure that made every engineer more effective. This is where the economics shift. If AI removes 10–15% of the administrative and coordination work from an engineer’s week, their effective output goes up. The cost of hiring remains the same, but the value per hire increases. In that scenario, reducing headcount undercuts your own leverage. Zapier leaned into this logic. As Kordampalos puts it, “We are doubling down and hiring even more people because we want to boost our productivity by using AI.” AI didn’t replace engineers; it replaced the parts of engineering work that prevent engineers from doing their best engineering. The move from summarizing meetings to a full agent ecosystem
Reported result
The source reports that existing employees increased throughput without extending working hours. It claims AI removes 10–15% of administrative work, raising effective output per engineer. No specific failure rates or detailed metrics beyond these general claims are provided in the text.
Human role
The original post does not say how a person stayed involved.
Limitations
The source does not detail specific integration costs, error handling protocols, or the exact technical stack beyond 'Zapier Agents'. It assumes a culture of 'build the robot' and relies on Slack as the coordination layer. Human approval is implied for governance but not explicitly defined as a mandatory step for every message.
Try it
Act as an Engineering Enablement Lead. I want to implement an async standup automation workflow using Zapier Agents to reduce meeting overhead for my team of [replaceable: number] engineers. 1. **Context**: My team currently holds daily standups in [replaceable: tool, e.g., Slack]. I want to replace this with async updates summarized by an agent. 2. **Task**: Create a Zapier Agent workflow that pings each team member at [replaceable: time] for their updates, collects responses, and generates a summary for the team channel. 3. **Safeguards**: Before enabling this for the whole team, run a dry run with [replaceable: 2-3 pilot users]. Show me the raw inputs and the generated summary. I must approve the summary format and accuracy before we roll it out to the full team. 4. **Output**: Provide the step-by-step configuration guide for the Zapier Agent, the prompt used for summarization, and a checklist for the pilot review. 5. **Success Metric**: We will measure success by a reduction in synchronous meeting time and positive feedback from pilot users on the clarity of the summaries. If any inputs are missing, ask me for clarification before proceeding.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.