Jul 1, 2026 · 26m · startup-ideas
AI Agents are the new SaaS
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Greg Isenberg presents a comprehensive masterclass on launching AI agent startups, detailing how entrepreneurs can transition from selling software tools to selling completed labor. The video outlines practical frameworks for workflow selection, agent architecture, customer pilots, SaaS wrappers, and a structured 30-day execution roadmap.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Greg holds 100% of the talking time here. How this is scored →
Greg as informed peer 0.0 Guest teaching 0.0 Guest disagreement 0.0 Greg pushing back 0.0
speaking balance: gold is Greg, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 22:06 Brief acknowledging interjection
In an entirely solo monologue episode, the only other speaker voice simply utters a single agreeable word ('Ok.').
Hardest push from Greg ▶ 9:35 Pushing back against overambitious Twitter AI demosThe host criticizes the common industry framing that agents must immediately be fully autonomous digital employees, arguing instead for bounded starter agents.
Biggest teaching moment ▶ 7:25 Educating founders on workflow nuanceThe host lectures the audience on how prompt engineers miss operational edge cases by failing to shadow real workers.
Greg holds their own ▶ 11:10 Synthesizing AI engineering principlesThe host demonstrates domain expertise by referencing Anthropic's guidance on earning autonomy through predictable deterministic workflows first.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Greg as informed peer | Guest teaching | Guest disagreement | Greg pushing back | Why |
|---|---|---|---|---|---|---|
| The Core Mental Model: The Product Is the Job | 0 | 0 | 0 | 0 | This is a solo monologue segment where the host outlines the core mental model that agent SaaS sells labor rather than traditional software tools. Because there is no guest interaction, all dynamic scores are set to zero. | |
| Identifying High-Value Workflows with Attached Paychecks | 0 | 0 | 0 | 0 | Host delivers a monologue on the five key traits of viable agent workflows with existing paychecks attached. No guest dialogue occurs. | |
| Shadowing Operators and Specifying Agent Architecture | 0 | 0 | 0 | 0 | Host explains the methodology of shadowing human operators before writing code or prompts in a solo presentation. There is no conversational pushback or schooling. | |
| Building the Minimal Useful Agent and Scaling Autonomy | 0 | 0 | 0 | 0 | Host describes the minimal useful agent framework and laddering autonomy from workflows to decisions. Pure solo instruction with no dialogue. | |
| Creating the SaaS Wrapper and Building Trust with Evals | 0 | 0 | 0 | 0 | Host details the importance of the product wrapper, dashboards, and eval test sets in a monologue format. No interaction occurs. | |
| Selling Initial Pilots as Labor and Productizing Patterns | 0 | 0 | 0 | 0 | Host outlines the strategy of running manual pilots and pricing them like labor before productizing repeatable workflows. Pure monologue. | |
| Distribution Tactics and Workflow Teardown Marketing | 0 | 0 | 0 | 0 | Host breaks down content marketing and workflow teardown distribution tactics solo. No debate or guest participation. | |
| The 30-Day Launch Plan and Concluding Takeaways | 0 | 0 | 0 | 0 | Host walks through a 30-day launch schedule with a brief one-word interjection from an uncredited speaker ('Ok.'), remaining entirely instructional. |
Statements from this episode (16)
Isenberg: AI agent TAM is larger than SaaS because it replaces labor
“Labor is a multi-trillion dollar market, right? So I think why people say agents is bigger than SaaS is because the total addressable market for agents is just way bigger. You know, it's human capital.”
Isenberg: Traditional SaaS sells tools while agent SaaS sells completed work
“SaaS sells software. Agent SaaS sells work. A normal SaaS product says, here is a tool a team could use basically. But an agent SaaS product will say, here is a job your team no longer has to do by hand.”
Isenberg: AI agent startups should replace annoying jobs cheaper than headcount
“And so when you're thinking about coming up with an idea for a startup here, think about, I handle this one annoying job better than a junior employee, faster than an agency, and it's cheaper than adding headcount.”
Isenberg: Target AI agents at workflows with existing payroll or agency spend
“If people are already paying for the work, they're paying an employee, they're paying an agency, a receptionist, a coordinator, dispatcher, they, there's an opportunity to sell that service cheaper And then unload some of that work that that person is doing to…”
Isenberg: Agent sweet spot is repetitive work requiring moderate human judgment
“So if a workflow is too basic basically the problem is basic automation, you know, zaps and stuff like that can do it. If it's pure human judgment the first version will break. So the sweet spot is repetitive work with enough judgment that AI can help.”
Isenberg: Founders must shadow human workers before coding AI agents
“You actually want to shadow the human before you build. So once you find the job, right, you, you've identified what the job is before you like prompt, before you start coding, just shadow a human being who does the job. I know a lot of founders want to skip t…”
Isenberg: AI agent specs require seven core components
“When you're specking out your agent, you know, I think it should have seven key parts. What wakes the agent up? What context does it need? What tools can it use? What is it allowed to do itself? Where does it need approval? When should it escalate and bring a …”
Isenberg: Building fully autonomous AI employees creates bad businesses
“Most people hear agent and imagine a fully autonomous employee. And that's how you get these demos that you see on Twitter and they don't really work and it ends up being a bad business.”
Isenberg: Agent founders should start with workflows and earn autonomy
“Many agent problems should start actually as workflows. So a workflow follows a predictable path. An agent decides more dynamically. Founders should earn autonomy by starting with the predictable path and adding judgment only when it creates value.”
Isenberg: In agent-first SaaS, the software wrapper creates customer trust
“And I think what separates a cool automation from a real Agent first SaaS product is the agent does the work, but the wrapper creates the trust.”
Isenberg: Using customer data for AI evaluations creates effective sales assets
“It's also like low key, a really good sales asset, because imagine telling a property manager, You know, we tested this on, you know, 50 of your old maintenance requests. It routed 42 correctly, flagged six of them for human review, and made two mistakes. Here…”
Isenberg: Fastest path to agent products is doing manual AI labor first
“So the fastest path is usually a pilot where you manually do the work with AI, and then you productize the repeated parts.”
Isenberg: Outcome pricing is the future business model for agent-first SaaS
“I really believe that outcome pricing is the future of how a lot of these agent first businesses and software is going to be priced.”
Isenberg: Workflow teardowns are the most effective marketing for AI agents
“Workflow teardowns is what I'm seeing in real time that's working really, really well.”