At a glance
WHAT IT’S REALLY ABOUT
Zapier CEO reveals practical AI workflows for culture and hiring.
- Wade Foster argues leaders can’t delegate “AI adoption” via memos; they must create hands-on space (hackathons, show-and-tells) so fear drops and practical usage rises.
- He demonstrates how meeting transcript data (via Granola) can be prompted into an “unspoken culture handbook,” then converted into concrete rubrics for hiring, performance, and communication.
- Foster walks through a Zapier Agents workflow that auto-evaluates interview transcripts against a job description and Zapier values, producing a yes/no/maybe recommendation and serving as a bias check.
- Finally, he shows using Grok/X as a sourcing tool to find “diamonds in the rough” creators and niche talent outside standard LinkedIn pipelines—while noting bot/noise and query-tuning tradeoffs.
IDEAS WORTH REMEMBERING
5 ideasLeaders must drive AI adoption with structured experimentation, not delegation.
Foster warns that pushing AI down the org via a memo leaves one IC “figuring it out for the whole company.” Creating play space (hackathons/show-and-tells) helps teams build familiarity and reduces fear through real tool use.
Make AI fluency measurable to change behavior.
Rubrics clarify what “good” looks like at different levels (e.g., PM competencies) and shift incentives toward adoption. What gets measured and rewarded becomes the default path for skill-building.
Meeting transcripts are a high-signal dataset for discovering real company culture.
Using Granola “recipes,” Foster generates an “unspoken culture handbook” from accumulated meetings—often more specific than officially written values. This output can be stress-tested against stated principles to find alignment and gaps.
Turn culture outputs into operational artifacts: interview prompts, JDs, and performance criteria.
Once culture is expressed as concrete behaviors (do/don’t examples), it becomes usable in hiring and evaluation. Foster suggests piping the “unspoken culture” into ChatGPT to generate scoring prompts for interviewers.
Interview-evaluation agents can serve as bias checks and consistency layers.
Zapier Agents can ingest interview transcripts + job description + values rubric, then send a structured yes/no/maybe recommendation with reasoning. Foster uses it as a thought partner—especially when interviewing outside his core domain expertise.
WORDS WORTH SAVING
5 quotesI see a lot of CEOs fall to the delegation trap. They write the AI memo… and then they don't do anything else.
— Wade Foster
Once people put their hands on the tools, I find that some of the fear goes away.
— Wade Foster
AI is this infinitely patient coach.
— Wade Foster
There is so many tasks that are not economically valuable right now… and these are the areas where AI and agents really thrive.
— Wade Foster
If you've ever seen standard operating procedures… you've seen an agent.
— Wade Foster
High quality AI-generated summary created from speaker-labeled transcript.
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome