

163 episodes
by Greg Isenberg · English
Get your creative juices flowing with The Startup Ideas Podcast. Published twice a week, we bring you free startup ideas to inspire your next venture. Hosted by Greg Isenberg, CEO of Late Checkout and former advisor to Reddit and TikTok. Subscribe so you don't miss out. For more startup ideas, we created a database of 30+ startup ideas you can take at https://gregisenberg.com/30startupideas

The host reviews five (described as “six” earlier, but only five detailed) free/open-source GitHub repos that help in the “agentic era,” emphasizing practical workflows: making AI writing sound…
2 Sep 2026 · 25m · 7 chapters
The “marketing engineering” era—using AI agents plus data, code, and taste to turn market signals into pipeline, faster experiments, sharper positioning, and a continuously learning marketing system.
31 Aug 2026 · 35m · 13 chapters
WebMCP (Google/Microsoft experimental) and how websites can expose “agent-ready” actions so AI agents can complete tasks in-browser without scraping the full DOM.
26 Aug 2026 · 29m · 11 chapters
How to make startups “AI-native” (LLM-readability, role-based AI loops, human-in-the-loop), plus startup takeaways from iMessage agent directories, customer support as a growth loop, and broader…
25 Aug 2026 · 41m · 13 chapters
How to use GrokBot to run a “1-person company” where a small set of AI agents executes an entire business end-to-end (research, content, sales, automation), with best practices to avoid “agent creep”…
21 Aug 2026 · 44m · 20 chapters
“Skillsmaxxing” for multiplayer AI—using “skills” (SOPs for Claude/Codex) to standardize repeatable workflows across a team, and sharing them via GitHub-backed plugins with version control and…
19 Aug 2026 · 32m · 21 chapters
How to configure Anthropic Claude Code to behave like “AI employees” by setting up a repo-based operating system (workspace, memory, briefs/plan mode, specific tickets, “eyes” via desktop…
17 Aug 2026 · 48m · 18 chapters
How to build an “AI-native” company/workforce in 2026 using proactive agent teams, shifting from “managing agents” to enabling them, and designing the right context, goals, and tool access.
12 Aug 2026 · 48m · 24 chapters
How Cloudflare’s AI-agent monetization (AI crawl control, pay-per-crawl, and X402/HTTP 402 payment-required) could create a new “agent web” business model where agents pay for useful resources…
10 Aug 2026 · 34m · 18 chapters
Building AI “marketing agents” that generate leads and run outbound on autopilot, plus an organic variant that turns internal conversations into scheduled LinkedIn content.
5 Aug 2026 · 44m · 20 chapters
Explains “graph engineering” as a practical way to structure AI work into a workflow graph (planner → parallel researchers → skeptic/checker → merger/synthesizer → human approval), instead of one…
3 Aug 2026 · 26m · 9 chapters
Jack Dorsey’s new Buzz app, positioned as an “agentic Slack killer,” where teams chat with software agents that can brainstorm, code, run workflows, and integrate with tools and APIs.
28 Jul 2026 · 39m · 15 chapters
How to deploy “marketing agents” using Claude Code plus MCP-style tool integrations to run marketing in the cloud, with a focus on Facebook Ads feedback loops and data-driven creative iteration.
27 Jul 2026 · 38m · 12 chapters
How to run teams of AI agents 24/7, focusing on cloud agents, automations, and building a “software factory” that writes, reviews, and ships code continuously.
24 Jul 2026 · 45m · 19 chapters
Forward Deployed Engineering (FDE) for AI—why it’s valuable now that “intelligence” is commoditized, what FDEs do, and a claimed 30-day roadmap to become an FDE.
20 Jul 2026 · 52m · 21 chapters
A “$999 AI tools assessment” for small businesses (2–20 employees; $500K–$5M revenue) that identifies time drains via a 45-minute discovery call, then prescribes 3–7 off-the-shelf AI/SaaS tools to…
15 Jul 2026 · 1h 1m · 25 chapters
How to use “loop engineering” (build/verify/measure/learn with AI agents) to automate business growth, especially SEO, plus examples for ads, product feedback, and other marketing loops.
13 Jul 2026 · 40m · 18 chapters
Grok 4.5 as a major upgrade for AI agents running inside Hermes/OpenClaw—positioned as an “AI co-founder” that can access tools, run tasks quickly, and help build and market startups (websites,…
10 Jul 2026 · 56m · 22 chapters
The hosts test OpenAI’s GPT 5.6 inside Codex for a month, arguing it can act like an “operating system” for knowledge work and business automation.
9 Jul 2026 · 49m · 22 chapters
The episode argues “AI agents are the new SaaS,” positioning agent-first startups as selling “work” (jobs customers no longer do manually) rather than software tools.
1 Jul 2026 · 26m · 10 chapters
The host argues “learn AI” is bad advice and instead outlines six practical skill sets for the “agentic era,” with concrete “first reps” for each.
25 Jun 2026 · 30m · 12 chapters
GLM 5.2 for local AI—why it’s outperforming on benchmarks, how to set it up in Cursor/Codex via OpenRouter, and how to chain models to offset GLM 5.2’s lack of vision.
23 Jun 2026 · 23m · 11 chapters
How to build and monetize consumer iOS apps using AI (“vibe coding”), with a framework to reach ~$10K/month via a strong app idea, a single “gotcha” feature, and distribution through Instagram/TikTok…
15 Jun 2026 · 48m · 24 chapters
The episode argues that “frontier” AI like Claude Fable 5 can be banned or priced out because users rent access, so builders should adopt local models that run entirely on their own hardware.
13 Jun 2026 · 25m · 10 chapters