How I AI: Zapier CEO Wade Foster's Playbook for AI-Powered Recruiting and Culture Building

I was so excited to sit down with Wade Foster, the co-founder and CEO of Zapier. In my opinion, Zapier has done one of the most exceptional jobs of not just adding AI features to their product, but fundamentally rethinking how an entire company can operate in the age of AI.

So many leaders fall into what Wade calls the "delegation trap." They write the big AI memo, tell their teams to go figure it out, and then step back. Wade does the opposite. He believes leaders need to be in the trenches, running hackathons, doing show-and-tells, and creating space for their teams to play and learn. He lives by one of Zapier's core values: "Don't be a robot, build a robot."

In this episode, Wade pops open his screen and walks us through a complete, end-to-end playbook for how he, as CEO, uses AI to shape culture, hire great people, and drive the company forward. We’re not just talking about theory; we're looking at the exact tools, prompts, and workflows he uses every day. We'll explore how he turns unstructured meeting data into a clear cultural handbook, builds an AI agent to act as a co-interviewer, and even uses Grok to find diamonds in the rough that traditional recruiters would never find. Let’s get to it.

Workflow 1: Extracting Your "Unspoken Company Culture" with Granola

Every CEO talks about culture, but how do you measure it? How do you know if the values written on the wall match the reality of how your team operates day-to-day? Wade showed me a brilliant workflow for turning the vast, unstructured data of company meetings into a concrete, data-driven understanding of company culture.

Context: Moving from Stated Values to Real-World Behavior

Many companies have a values document, like Zapier's culture and values page, that outlines ideals like "Default to action." But what does that actually look like in practice? To create effective hiring rubrics and performance standards, you need specific examples of good and bad behavior. This is where AI can turn ambient conversation into a structured asset.

Step-by-Step: Using Granola's Recipe Feature

Wade uses Granola, a meeting recorder and summarizer, which is running in all of his meetings. This creates a massive corpus of data about how decisions are made, how people communicate, and what behaviors are implicitly rewarded. Here's how he leverages it:

1. Capture Everything: Let a tool like Granola run in the background of your meetings for an extended period. The more data it has, the more accurate the cultural analysis will be.
2. Use the "Unspoken Culture" Recipe: Granola has a built-in feature called "Recipes," which are essentially pre-built prompts. Wade runs a recipe called "Build the unspoken company culture handbook."
3. Analyze the Output: The AI analyzes all the meeting transcripts and generates a document that describes how the organization actually works, based on observed behaviors. Wade mentioned he was shocked at how well it captured specifics that even he hadn't fully articulated before.

When I first ran this, I was shocked. We spent a lot of time thinking about our culture, writing about our culture... but as I read through it, I was like, wow, this actually gets at the specifics in a way that even I hadn't figured out quite how to do.

From Culture Doc to Hiring Rubric

This generated document is incredibly powerful. It's not just an interesting reflection; it's a practical tool. Wade takes this output and feeds it into a tool like ChatGPT with a prompt like:

"Hey, can you take this unspoken culture and actually generate a set of scoring prompts for how to evaluate somebody in an interview against these traits that match Zapier?"

This closes the loop, turning ambient meeting data into a concrete, actionable rubric for hiring, performance reviews, and setting clear expectations across the company. It's a fantastic example of using AI to do the CEO's job—being the carrier of culture—better.

Workflow 2: Building an AI Interview Evaluation Agent

Once you have a clear rubric, the next challenge is applying it consistently and without bias during the hiring process. Wade, who interviews candidates across many different disciplines, showed me how he built an AI agent to act as his thought partner and bias-checker.

Context: Scaling High-Quality, Unbiased Interview Feedback

As a CEO, it's impossible to be a deep expert in every role you hire for. It's also easy for personal biases or the flow of conversation to cause you to miss key evaluation points. Wade built an agent using Zapier Agents to provide a consistent, objective second opinion on every interview he conducts.

Step-by-Step: Creating the Interview Agent

This agent is surprisingly simple to set up but incredibly effective. Here’s the breakdown:

1. Set the Trigger: The agent kicks off whenever a new meeting note from Granola is added to a specific folder, in this case, his "New Interviews" folder.
2. Provide Instructions: This is the core of the agent. Wade gave it a clear persona and a multi-step task. Here are the instructions he shared:

You are an expert hiring evaluator at Zapier. Your task is to review the interview transcript and notes provided by Granola. You're reviewing the job description provided as a knowledge source and Zapier's company values provided as a knowledge source to determine whether a candidate should advance in the hiring process. You want to evaluate the candidate's functional expertise, their values alignment... your goal is to recommend yes, no, or maybe to this candidate and provide your reasoning. Give me three to five sentences on why you are recommending this. Then go ahead and email me the evaluation.

1. Upload Knowledge Sources: The agent's analysis is grounded in specific context. Wade uploaded two Google Docs as knowledge sources:

• The Zapier Values Rubric (derived from Workflow 1)
• The specific Job Description for the role (e.g., Social Media Specialist)

1. Define the Action: The final step is for the agent to send Wade an email with its complete evaluation.

Enhancing the Agent with Copilot

During our session, I offered Wade two suggestions to make the agent even better, inspired by a workflow from Zach Davis at LaunchDarkly:

• Evaluate the Interviewer: Add a section that gives feedback on the interviewer's performance against the rubric (e.g., "You forgot to ask about X.").
• Surface the Decision: Put the hire decision (YES/NO/MAYBE) directly in the email subject line for faster triaging.

Wade immediately used the built-in Copilot within Zapier Agents to incorporate these changes live. He simply typed the suggestions in natural language, and Copilot updated the agent's instructions in seconds. This shows how quickly you can iterate and improve these AI systems once they're in place.

The Result: A Consistent, Actionable Evaluation

The output is a clean, structured email that gives a clear recommendation, supports it with reasoning tied directly to the job description and company values, and helps Wade make better, faster hiring decisions. It acts as a perfect thought partner, catching things he might have missed and ensuring every candidate gets a fair and consistent evaluation.

Workflow 3: Sourcing "Diamonds in the Rough" with Grok

This last workflow was a complete surprise and a first for How I AI. While most recruiters live on LinkedIn, Wade has found a creative way to use Grok—the AI with real-time access to X—to find passionate, under-the-radar talent.

Context: Finding Chronically Online Talent

For a role like a social media manager, you want someone who is truly a native of the internet. These "diamonds in the rough" are often active in niche communities, sharing tutorials and building a following, but might not have a polished LinkedIn profile. Grok is uniquely positioned to find these people.

Step-by-Step: Prompting Grok for Candidates

Wade's process is an iterative conversation with Grok to zero in on the right profiles.

Start with an Open-Ended Prompt: He began with a broad query to find relevant people on X.

help me find posters on X that are fans of Zapier, no code, agent building, automation, and related topics. I want posters that share tutorials and education related ideas. These posters should have modest followings... I'm looking for diamonds in the rough... We are on a budget, so look for folks outside the Bay Area. Give me 10 ideas.

Analyze and Refine: The first set of results included some promising leads but also some potential bots. Wade refined his prompt in real-time:

• let's do not a bot
• give me people with real faces as for profile photos
• let's do United States located folks

Pivot to Other Platforms: When he wanted to find video creators, he simply pivoted his request:

how about finding 10 YouTubers

The Outcome: A New Channel for Talent

This workflow is a numbers game. Not every result is a winner, but it surfaces people and even geographic hotspots of community interest (like a no-code scene in Bangalore) that would be completely invisible through traditional sourcing methods. It’s a perfect example of using the unique strengths of a specific model (Grok's real-time social access) to gain a competitive edge in recruiting. Wade has even used this same technique to source highly specialized technical AI talent, another community that is extremely active on X.

Conclusion: The AI-Native CEO

Wade's three workflows provide a powerful, interconnected playbook for any leader looking to do more than just talk about AI. He starts by using AI to get a data-driven pulse on the company's real culture. He then uses that cultural blueprint to build AI systems that help him hire people who will thrive in that environment. Finally, he uses novel AI tools to find talented people in places his competitors aren't even looking.

This is what it means to lead AI adoption from the front. It’s not about writing memos; it’s about rolling up your sleeves, using the tools yourself, and reimagining core business functions like hiring and culture-building from the ground up. By automating tedious tasks and creating systems for feedback and evaluation, you free up human time and energy to focus on what really matters: finding, hiring, and inspiring the best people. I hope this gives you some practical ideas for how you can start building your own AI-native workflows today.

Thanks to Our Sponsor

This episode is brought to you by Brex—The intelligent finance platform built for founders.

Episode Links

• YouTube: https://youtu.be/ZKnASs_d7aE?si=QPjVtcLqyaah9ldk
• Spotify: https://open.spotify.com/episode/4iMakw8FwEgWe5Q2iNp6HZ?si=EnQNr8VzTEicy2AwTCJFNw
• Apple Podcasts: https://podcasts.apple.com/us/podcast/how-i-ai/id1809663079?i=1000743802917