Supra Insider

Chris Gomes set out to hire four AI product managers. Seven months later, his biggest lesson was not about AI at all.
In this episode of Supra Insider, Marc Baselga and Ben Erez sit down with Chris Gomes, VP of Product at Conveyor, the Series B startup that automates responses to security questionnaires and RFPs. Chris walks through the seven months he spent hiring four AI PMs, why he spent so long defining what an “AI product manager” even means, and the realization that culture fit, not AI skill, was the thing that actually predicted success.
They explore how he rebuilt a stalled interview process by pulling the most important screens to the front, the MOC framework he uses to map each interview step to specific competencies, his case for work trials and customer role-plays, how he treats references and back-channels, and why the gut-level question of whether you’d enjoy working with someone deserves more weight than most rubrics give it.
If you’re a hiring manager trying to run a tighter, higher-signal process, a founder thinking about your first product hires, or a PM preparing for interviews and wondering what teams are really evaluating, this episode is for you.
All episodes of the podcast are also available on Spotify, Apple and YouTube.
New to the pod? Subscribe below to get the next episode in your inbox 👇
Want company-specific interview intel?
If you’re preparing for PM interviews at Meta, Anthropic, OpenAI, Google, Stripe, DoorDash, Uber, or Figma… we think you’re going to love Insider Loops. You can read about it in Ben’s LI post, Marc’s LI post or check out the site directly below.
Supra Insider listeners get $50 off all interview guides with promo code SUPRAINSIDER50
In this episode, we covered the following topics:
* (03:47) Chris sets the scene: hiring four AI PMs at Conveyor after their Series B
* (05:09) The big realization: culture fit matters far more than AI skills, and most of the AI part is teachable
* (06:42) How to actually interview for culture by naming what’s genuinely different about yours
* (08:25) Conveyor’s specifics: hungry, humble, smart, plus customer discovery and the honesty-versus-decisiveness balance
* (12:33) Why they were whiffing, and bringing in recruiter Chris Lee to fix the process
* (13:35) Treating hiring like a product: pull the most important screens forward and de-risk the riskiest parts first
* (19:44) Respecting candidates’ time by cutting performative steps that generate mixed signal
* (21:49) Closing with transparency: walking finalists through the board deck, good, bad, and ugly
* (24:16) The work-trial dream, and a more interactive take-home as a middle ground
* (30:27) Their take-home plus a live customer role-play that proved the most predictive step
* (35:56) The MOC framework: mission, outcomes, and stack-ranked competencies mapped to each step
* (40:23) Taking references seriously, the “Who” method, and where back-channels actually help
* (51:02) The vibes question: can you capture whether you’d enjoy working with someone
* (59:30) Meeting finalists in person, even for remote roles, and the jerk-to-the-waiter signal
* (1:03:38) How much to share with candidates upfront, and valuing thinking on your feet
* (1:11:08) Evaluating AI depth without chasing the flavor of the day, plus what they under-tested
* (1:18:53) Gratitude Corner: Chris thanks Chris Lee, Sydney, and Chas
* And more!
Links:
* Chris Gomes on hiring AI PMs:
* https://www.linkedin.com/posts/gomeschris_over-the-past-7-months-ive-hired-4-product-share-7401947584958787584-O9Um/
* https://www.linkedin.com/posts/gomeschris_takeaway-2-from-hiring-four-pms-w-ai-depth-share-7440043809100484608-ziTH/
* Conveyor: https://www.conveyor.com/
* Chris Lee: https://www.linkedin.com/in/thechrislee/
* Sydney Shea: https://www.linkedin.com/in/sydneyfshea/
* Chas Ballew: https://www.linkedin.com/in/chasballew/
* Chris Gomes: https://www.linkedin.com/in/gomeschris/
* Marc: https://www.linkedin.com/in/marcbaselga/
* Ben: https://www.linkedin.com/in/benerez/
If you’re a new subscriber, we encourage you to check out some of the recent episodes you might have missed:
To support the podcast, please check out the links below:
* Check out Ben’s AI Practice Copilot and top-rated Maven course: Self-Paced PM Interview Bootcamp: Product Sense & Analytical Thinking (selected by Lenny Rachitsky as a top course 🔥). Get 25% off the course with promo code suprainsider (auto-applied here).
* Supra has teamed up with Maven to bring you something special – courses that our own community members have personally curated. And because you’re part of the Supra family, you can get up to 35% in some of these handpicked selections with code SUPRAxMAVEN.



This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit suprainsider.substack.com

This episode contains some screen sharing so it’s best watched on YouTube
What happens when one product leader decides to stop copy-pasting between chat windows and instead build an operating layer that puts coding agents in the hands of an entire company?
In this episode of Supra Insider, Marc Baselga and Ben Erez sit down with Kyler Ross, Head of Product at Cloaked, to walk through the internal “harness” he started building last Thanksgiving: an agent-friendly system of context files and scripts that lets agents read from and write to the team’s real tools. Kyler explains how it gets installed on every company machine, why he treats each new agent session like onboarding an employee, and how a self-improving loop of skills and automated reviews keeps it getting better.
They explore his day-to-day setup for running many agents at once, why worktrees and Claude Code hooks exist to make failure nearly impossible, a one-on-one prep skill that pulls context from every corner of the company, and the layered guardrails, including a nightly “librarian” agent, that keep confidential information from leaking.
If you’re a product or engineering leader trying to make your team more AI-native, someone wiring agents into real workflows, or anyone wrestling with how to run agents safely at scale, this episode is for you.
All episodes of the podcast are also available on Spotify, Apple and YouTube.
New to the pod? Subscribe below to get the next episode in your inbox 👇
Want company-specific interview intel?
If you’re preparing for PM interviews at Meta, Anthropic, OpenAI, Google, Stripe, DoorDash, Uber, or Figma… we think you’re going to love Insider Loops. You can read about it in Ben’s LI post, Marc’s LI post or check out the site directly below.
Supra Insider listeners get $50 off all interview guides with promo code SUPRAINSIDER50
In this episode, we covered the following topics:
* (01:15) The Thanksgiving origin: copy-pasting between chat windows and losing prompts pushed Kyler to build a harness
* (02:32) What the harness is: an agent-friendly directory of context plus scripts that give the agent hands to read and write
* (06:18) How Cloaked deploys it: installed by default on every company machine, with nearly everyone using it
* (07:07) Why he owns ~80% of the code, treating it as a nights-and-weekends hobby separate from his day job
* (08:52) The self-improving loop: anyone can trigger a skill to fix a bad session and submit a PR
* (12:23) Using the harness as a testing ground for a more AI-native way to ship code
* (15:17) Why Slack agents, not Claude Code for everyone, democratize access for non-technical roles
* (16:21) Branding it “PMAI” and the Start Cloaked app that onboards employees
* (30:04) His iTerm2 setup and a keyboard shortcut that launches Claude Code in the right place
* (35:48) Maximizing concurrent agents, and the “infinite canvas” he wishes existed
* (39:14) Why running many agents at once becomes a key skill differentiator, and its fatigue tax
* (44:33) Worktrees and Claude Code hooks as deterministic guardrails that make failure nearly impossible
* (50:44) A one-on-one prep skill that pulls Slack, Granola notes, tickets, and docs into one place
* (54:53) Layered confidentiality guardrails and the nightly “librarian” agent, plus the 50,000-line PR story
* (1:01:20) Gratitude Corner: Kyler thanks Mike Carafa and Aashima Ratti
* And more!
Links:
* Cloaked: https://www.cloaked.com/
* Cloaked Job Board: https://jobs.ashbyhq.com/cloaked
* Sr. Product Manager opening: https://jobs.ashbyhq.com/cloaked/f1b12f94-a114-4d4b-a78e-5a6a2b6ee1fe
* Mike Carafa: https://www.linkedin.com/in/michaelcarafa/
* Aashima Ratti: https://www.linkedin.com/in/akratti/
* Kyler Ross: https://www.linkedin.com/in/kyler-ross-63552595/
* Marc: https://www.linkedin.com/in/marcbaselga/
* Ben: https://www.linkedin.com/in/benerez/
If you’re a new subscriber, we encourage you to check out some of the recent episodes you might have missed:
To support the podcast, please check out the links below:
* Supra has teamed up with Maven to bring you something special – courses that our own community members have personally curated. And because you’re part of the Supra family, you can get up to 35% in some of these handpicked selections with code SUPRAxMAVEN.
* Check out Ben’s AI Practice Copilot and top-rated Maven course: Self-Paced PM Interview Bootcamp: Product Sense & Analytical Thinking (selected by Lenny Rachitsky as a top course 🔥). Get 25% off the course with promo code suprainsider (auto-applied here).



This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit suprainsider.substack.com

What does it take to walk away from a decade in product, and a job most people would envy, to bet on yourself?
In this episode of Supra Insider, Marc Baselga and Ben Erez sit down with Peter Yang, who just left his product lead role at Roblox to go full-time on his newsletter and podcast, Behind the Craft and build his own projects. Peter talks through the trade-offs of solopreneur life, why his calendar is suddenly empty, and how he uses an AI personal advisor with three principles to decide what to say no to.
They explore his day-to-day AI builder stack, from running Codex as a daily driver to using Hermes for his recurring scheduled tasks, his working definition of slop and why he guards against it, and what he’s actually measuring as success now that nobody is handing him a promotion.
If you’re a PM weighing whether to leave a stable job to build on your own, a creator trying to scale output without sliding into slop, or anyone wiring AI agents into their daily work, this episode is for you.
All episodes of the podcast are also available on Spotify, Apple and YouTube.
New to the pod? Subscribe below to get the next episode in your inbox 👇
Want company-specific interview intel?
If you’re preparing for PM interviews at Meta, Anthropic, OpenAI, Google, Stripe, DoorDash, Uber, or Figma… we think you’re going to love Insider Loops. You can read about it in Ben’s LI post, Marc’s LI post or check out the site directly below.
Supra Insider listeners get $50 off all interview guides with promo code SUPRAINSIDER50
In this episode, we covered the following topics:
* (01:21) Ben opens on the “unstructured abyss” of solopreneur life and what Peter’s most excited to build
* (05:56) Peter’s AI “personal advisor” and the three principles he uses to decide what to say no to
* (07:24) Choosing the identity of “builder” and building in public over polished talking-head content
* (21:38) How it started: bored at a fintech job, a four-hour OKR meeting, and his first Substack post
* (24:35) Why he left: not wanting the VP life of all-day meetings and a game he didn’t want to win
* (30:10) Accountability rituals: a content cadence plus AI-generated weekly business and health reports
* (32:03) Using AI to spot growth trends he’d have missed, like Substack notes and LinkedIn link posts
* (33:36) Codex as his daily driver and running cron jobs using Hermes
* (42:18) Symlinking skills so Codex and Claude Code share the same setup
* (47:02) Fighting bloat with a “skill that edits skills” and the kill-a-feature problem for AI systems
* (52:24) Defining slop and why he guards craft, with the “would my kids see this” test
* (58:54) What success looks like: around $1M a year and full control over his time
* (1:02:42) Treating a graveyard of dead vibe-coded projects as R&D for the thing with product-market fit
* (1:11:32) Where to find Peter and how to help, plus the Gratitude Corner thanking his wife and Ben
* And more!
Links:
* Peter Yang’s recent post: https://open.substack.com/pub/peteryang/p/i-quit-my-high-paying-product-job-to-bet-on-myself
* Peter Yang’s X: https://x.com/petergyang
* Peter Yang’s Substack: https://creatoreconomy.so/
* Peter Yang: https://www.linkedin.com/in/petergyang
* Marc: https://www.linkedin.com/in/marcbaselga/
* Ben: https://www.linkedin.com/in/benerez/
If you’re a new subscriber, we encourage you to check out some of the recent episodes you might have missed:
To support the podcast, please check out the links below:
* Check out Ben’s AI Practice Copilot and top-rated Maven course: Self-Paced PM Interview Bootcamp: Product Sense & Analytical Thinking (selected by Lenny Rachitsky as a top course 🔥). Get 25% off the course with promo code suprainsider (auto-applied here).
* Supra has teamed up with Maven to bring you something special – courses that our own community members have personally curated. And because you’re part of the Supra family, you can get up to 35% in some of these handpicked selections with code SUPRAxMAVEN.



This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit suprainsider.substack.com

What does it take to bring AI into businesses that run on physical work, human judgment, and processes nobody has ever written down?
In this episode of Supra Insider, Marc Baselga and Ben Erez sit down with Noah Levin, founder of Serious People, to unpack what he calls being a “free-range AI consultant.” Noah explains why most of his work is business consulting from first principles rather than AI consulting, why agents still need humans to deliver real value, and how he groups AI for any company into three buckets: a coworker, an operator, and a product or engineering capability.
They explore how AI is collapsing the distance between a conversation and a working prototype, why the new IP is business judgment instead of code, why he believes everything is becoming product management, and the humility it takes to solve problems on a client’s terms inside companies that aren’t, and shouldn’t be, run like tech startups.
If you’re a product leader figuring out where AI actually creates leverage, an operator weighing whether to go independent, or a builder realizing that distribution now matters more than the thing you build, this episode is for you.
All episodes of the podcast are also available on Spotify, Apple and YouTube.
New to the pod? Subscribe below to get the next episode in your inbox 👇
Want company-specific interview intel?
If you’re preparing for PM interviews at Meta, Anthropic, OpenAI, Google, Stripe, DoorDash, Uber, or Figma… we think you’re going to love Insider Loops. You can read about it in Ben’s LI post, Marc’s LI post or check out the site directly below.
Supra Insider listeners get $50 off all interview guides with promo code SUPRAINSIDER50
In this episode, we covered the following topics:
* (01:33) Noah defines the “free-range AI consultant”: applying AI with no fixed stack or playbook, adapting to whatever the business needs
* (02:32) Why most of the work is business consulting from first principles, with AI as a tool in service of the goals
* (09:55) Why this need exists at all: forward deployed engineer postings up 700%, and Dan Shipper’s point that every agent needs a human
* (16:20) The three buckets every company should think about: AI as coworker, AI as operator, and AI as product/engineering
* (19:05) The gym, Granola, and Claude Code story: a 20-minute conversation becomes a research report and a working prototype before the meeting ends
* (22:17) The floor went up, but raw model output isn’t a finished product; baking work into reusable skills so IP compounds
* (27:07) Why the new IP is business judgment, not code, and why you shouldn’t automate a process until you know it’s the right one
* (29:30) Why everything is becoming product management, and why PMs get the most leverage from AI
* (35:31) “Life is a series of prompts”: every conversation is a prompt for the next task
* (40:18) The optimization trap: the unopened Mac Mini and balancing system-building against actual leverage
* (48:25) Why not every company is a tech company: McDonald’s three-legged stool and the humility to solve on a client’s terms
* (1:04:33) Distribution over building: in a world where anyone can build anything, market choice matters more than the product
* (1:08:50) Why the real opportunity is making contact with the world outside tech
* (1:13:31) Where to find Noah, plus the Serious People Podcast launching in July
* (1:19:27) Gratitude Corner: Noah thanks his first boss, Rick Boultinghouse, from his Navy Pier days
* And more!
Links and Resources:
* Lenny’s podcast with Dan Shipper: https://www.youtube.com/watch?v=4D3hDmGhFhA
* Matt Van Horn: https://x.com/mvanhorn/articles
* Something big is happening: https://x.com/mattshumer_/status/2021256989876109403
* Something small is happening:
https://myriadperspectives.com/p/something-small-is-happening
* Serious People: https://seriouspeople.ai/
* Serious People AI Podcast: https://seriouspeople.ai/podcast
* Rick Boultinghouse: https://www.linkedin.com/in/rick-boultinghouse-435a813/
* Noah Levin: https://www.linkedin.com/in/noahlevin/
* Marc: https://www.linkedin.com/in/marcbaselga/
* Ben: https://www.linkedin.com/in/benerez/
If you’re a new subscriber, we encourage you to check out some of the recent episodes you might have missed:
To support the podcast, please check out the links below:
* Supra has teamed up with Maven to bring you something special – courses that our own community members have personally curated. And because you’re part of the Supra family, you can get up to 35% in some of these handpicked selections with code SUPRAxMAVEN.
* Check out Ben’s AI Practice Copilot and top-rated Maven course: Self-Paced PM Interview Bootcamp: Product Sense & Analytical Thinking (selected by Lenny Rachitsky as a top course 🔥). Get 25% off the course with promo code suprainsider (auto-applied here).



This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit suprainsider.substack.com

What if the thing holding you back from posting isn’t laziness or a lack of ideas, it’s that nobody ever told you the cringe feeling goes away, and what it actually looks like to build trust with an audience without burning it?
In this special episode of Supra Insider, Ben Erez sits down with Hilary Gridley, creator of the Maven course “How to Become a Supermanager with AI,” and Mallory Contois, former VP of Growth at Maven and founder of The Old Girls Club, for a candid conversation about self-promotion, audience building, and the surprisingly practical mechanics of showing up consistently online without losing yourself in the process. Hilary published a guest post on Lenny’s Newsletter the same morning this was recorded, and that pipeline, from writing to course to full-time career, is exactly what the conversation unpacks.
They cover why trust is the most durable professional asset you can build, how to think about the value exchange between creator and audience, why the psychology of “doing work in private and handing in the final product” makes content creation feel so unnatural, and what both of them actually do to stay consistent without spiraling into algorithm-chasing. Mallory drops a deceptively simple Apple Notes system for never running out of ideas. Hilary makes a sharp case for starting with a talk instead of a newsletter. And both of them are refreshingly honest about the fact that posting still feels mortifying sometimes, and why you should do it anyway.
If you’re a PM, operator, or founder who has been sitting on the sidelines of content creation because it feels cringe, trying to figure out which platform to start with, or building an audience and wondering how to grow it without compromising the thing that makes your voice worth following- this episode is for you.
All episodes of the podcast are also available on Spotify, Apple and YouTube.
New to the pod? Subscribe below to get the next episode in your inbox 👇
Want company-specific interview intel?
If you’re preparing for PM interviews at Meta, Anthropic, OpenAI, Google, Stripe, DoorDash, Uber, or Figma… we think you’re going to love Insider Loops. You can read about it in Ben’s LI post, Marc’s LI post or check out the site directly below.
In this episode, we covered the following topics:
* (03:35) Hilary’s background: from music blogging to Maven to going full-time after maternity leave
* (06:18) Mallory’s background: The Old Girls’ Club, being an 87% introvert, and experimenting without a grand plan
* (15:15) Why trust is the most durable professional asset, and how content creation builds it at scale
* (18:44) The value exchange principle: give more than you ask for, measured across a year not a single post
* (20:47) “You wanna be a magnet for the people you most want to work with” — Hilary on her core motivation
* (23:22) Overcoming cringe: why it still feels mortifying every time, and why that doesn’t go away
* (27:00) Mallory’s framework: why we were trained to do work in private and why that makes content feel unnatural
* (29:47) What happens when you post publicly: the community you find by being brave enough to start
* (31:26) Hilary on being misunderstood: simplify, accept the trade-off, and learn from the rude comments
* (36:31) Mallory’s Apple Notes system: the messy backlog that keeps content ideas authentic
* (39:12) Live Q&A: which platform to start with? Hilary on using Substack as hub and LinkedIn as distribution
* (43:06) “Please, for the good of society, change the composition of voices on these platforms”
* (48:00) Live Q&A: Hilary’s advice on starting with a talk as your first content format
* (49:01) Live Q&A: networking as a byproduct — content as a bat signal for the right people
* And more!
Links:
* Mallory’s Substack: https://www.readgoodwork.com/
* The Old Girls Club by Mallory: https://www.jointheogc.com/
* Building a Personal Brand (Course by Mallory): https://maven.com/mallory/building-a-personal-brand
* Hilary’s AI Guide: https://couchto5k.ai/
* Hilary’s Substack: https://hils.substack.com/
* How to Become a Supermanager with AI by Hilary: https://maven.com/hilary-gridley/ai-powered-people-management
* Mallory Contois: https://www.linkedin.com/in/mallorycontois/
* Hilary Gridley: https://www.linkedin.com/in/hilarygridley/
* Marc: https://www.linkedin.com/in/marcbaselga/
* Ben: https://www.linkedin.com/in/benerez/
If you’re a new subscriber, we encourage you to check out some of the recent episodes you might have missed:
To support the podcast, please check out the links below:
* Check out Ben’s AI Practice Copilot and top-rated Maven course: Self-Paced PM Interview Bootcamp: Product Sense & Analytical Thinking (selected by Lenny Rachitsky as a top course 🔥). Get 25% off the course with promo code suprainsider (auto-applied here).
* Supra has teamed up with Maven to bring you something special – courses that our own community members have personally curated. And because you’re part of the Supra family, you can get up to 35% in some of these handpicked selections with code SUPRAxMAVEN.


This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit suprainsider.substack.com