The Startup Ideas Podcast (SIP) 🧃
@startupideaspod • 49,707 subscribers
Get startup ideas and practical tutorials on AI tools to build your business. Host: @gregisenberg Available on Spotify, Apple and YT
Videos
I believe vibe manufacturing will create thousands of companies that make $1M or more a year. In 2024, Lovable and Bolt made anyone a vibe coder. Now the same shift happens for physical products, and OpenAI's Astra was the first model of this era. Before, a hardware product meant your team needed a full-time 3D designer, a mechanical engineer, and an electrical engineer. Now you pay $20 a month for a frontier model to walk you through all of it. The market for your product can be small. If it costs little to make, there is a business. Hardware looked too hard to me. I did not know it, so I did not try. Now I have a hardware desk on the way. Make one thing: a printable object, a CAD file, a Blender model, or a circuit board. You owe it to yourself.
The Startup Ideas Podcast (SIP) 🧃74,517 görüntüleme • 25 gün önce
Boris Cherny, the creator of Claude Code, shared his entire setup. He runs 5-10 Claudes in parallel. Half his coding happens from his phone. Here's his 3-part formula for better results: Use the smartest model available — Counterintuitive: it's actually cheaper — Smarter model = fewer tokens = lower total cost — "Once the plan is good, the code is good" Invest in your Claude MD — Plain text file. No special format. — Whole team contributes multiple times a week — Every mistake Claude makes gets added so it never happens again Give Claude a way to verify its own output — Let it run the code. Let it see the browser. — "Imagine you're a painter wearing a blindfold" — Same thing for an AI that can never check its work His morning routine: wake up, kick off 3 sessions from his phone, check in later. His workflow: start in plan mode → lock the plan → auto-accept edits → done. No fancy setup. No complex tooling. Just multiple Claudes, a good plan, and a shared knowledge base.
The Startup Ideas Podcast (SIP) 🧃560,564 görüntüleme • 7 ay önce
One of the best businesses to run with GrokBot is a directory. A directory is basically just curated data on a website. You pick a topic and you say, here are the best XYZ in this topic. It's a good $10K, $25K, $100K a month business, Here's the whole setup: 1) Pick the niche. Pickleball equipment. Board games that are good for parties. Something you already have taste in. 2) Build the site on Astro. Astro specifically is really good at directories. 3) Set up a research agent. Every day it investigates a new brand and writes a great page on that brand as Markdown. 4) It hands the page to the engineer, who posts it. 5) Rinse and repeat until you're getting thousands of visits a month. 6) Then add an agent that collects referral links and does outbound to get deals. Step 6 only works after step 5. Nobody gives you a deal when the site has no traffic. Once the directory is getting visits, you have a foundation. Put a niche newsletter on a subdomain and link people over to the directory with your referral link. Those warm leads become sponsors in the newsletter. And the whole SEO game is changing right now. People are going to ChatGPT and Google AI overviews to consume content, and platforms like that need directories. There's too much information out there, so curated is what's helpful. Pick the niche, let it write the pages, then go get the deals
The Startup Ideas Podcast (SIP) 🧃108,068 görüntüleme • 2 ay önce
How does an FDE map the processes to deploy AI in a real business? We use 3 sources. Source 1 is people. - Companies tell us: "this person's been at the firm for 20 years and they just handle it." - We hear this from Fortune 50 companies and from SMBs. - So in finance, you talk to the head of AP, AR, reconciliations, billing, banking, and FP&A. You make a "human API." Source 2 is the systems of record. - With real-time access to Salesforce for 3 to 4 weeks, you see what data comes in and how often people correct it. Source 3 is the documents in SharePoint, Drive, Notion, Slack, Teams, and Gmail. If you use the system data without the interviews, you miss half of the picture. The split is different for each company. A large company can have 10 years of Salesforce data. An SMB has most of it in people's heads. Choose your angle for each company before you build.
The Startup Ideas Podcast (SIP) 🧃35,097 görüntüleme • 28 gün önce
FDE is the hottest role in tech. Here's the 30-day plan to actually become one: Week 1: Build the loop - Agent looping, tool usage, guardrails - Context, memory, and an audit trail - Checkpoint: one working agent, one task. Week 2: Make it recover - JSON schemas, not freeform text - Failure modes for the 1,000 ways it breaks - "There's only one way something goes right. A thousand ways it goes wrong." Week 3: Make it measurable - Retry logic + a golden dataset for evals - Test cheaper models on sub-tasks - Score it on revenue, risk, cost Week 4: Defend it like an FDE - The pain points, the architecture, the economics - Then pitch it to real businesses Think: - Week 1 = build - Week 2 = survive - Week 3 = measure - Week 4 = sell That's the whole ladder.
The Startup Ideas Podcast (SIP) 🧃109,117 görüntüleme • 3 ay önce
Before we automate a workflow, we always audit the client's current process. This client used 17 steps to pay one invoice. For them, that was a simple workflow. The audit took 2 to 3 weeks of interviews and process mining. After the redesign: - 17 steps to 7 - 24 days of cycle time to 6 - 6 exception loops to 1 - 18% straight-through rate to 87% - $31 per invoice to $6 At 18%, most invoices went through an exception route. We fixed that with process reengineering. Agents were only one part of it. Show the client the full mess first. It breaks their heart, but they need it.
The Startup Ideas Podcast (SIP) 🧃28,181 görüntüleme • 24 gün önce
how to actually get customers for your vibe coded app: - let AI sell for you (MCP servers) - build 10,000 SEO pages once (programmatic SEO) - vibe code a free tool by lunch (it markets itself) - get cited by ChatGPT (answer engine optimization) - allow your users to create content for you (viral artifacts) - buy a 10K subscriber newsletter for $5K - turn one voice memo into 20 pieces of content Pick two. Start this week. Stop just vibe coding. Start getting customers.
The Startup Ideas Podcast (SIP) 🧃169,277 görüntüleme • 7 ay önce
I run 34 AI agents with a 3-word prompt. "do smart things" That's the full instruction. My AI chief of staff is Simon. Simon runs the org. Everything under him touches my meeting transcripts, email, calendar, Notion, Stripe, Supabase, GitHub, my 2026 goals, my business, my family. Several times a day I want something to look across all of that and just go. I ran an org of 100 people at AWS. The part I loved was empowering people and watching them blow past what they thought they could do. Then it hit me that I was functioning at the limit of my own imagination, and every single thing my agents worked on was still initially prompted by me. The catch: the width expanded, the risk tier didn't. They are not sending 100 emails on their own. I still check all emails. Widen the scope before you widen the permissions.
The Startup Ideas Podcast (SIP) 🧃54,841 görüntüleme • 2 ay önce
Grok 4.5 is ~1/10th the cost of Fable and 10–15x faster. You'd think that saves you money. It doesn't and that's the best part. An 8-hour job now takes 1 hour. So I don't go home early. I pack 8x the work into the same day. I pay for every lab. Codex, Claude, Perplexity, Grok. This is the one that made me unstoppable. $300/mo plan, 24% left, bought it the day it dropped And it's the best money I spend.
The Startup Ideas Podcast (SIP) 🧃74,100 görüntüleme • 3 ay önce
The "marketing engineer" will be one of the most valuable jobs in a company. These people will command $250K, $500K, and $1M/year. The best marketer changes with every major tech wave: 1) Traditional marketer: makes people care with story (print, radio) 2) Digital marketer: owns the new digital channels (SEO, PPC) 3) Growth hacker: lives in loops, PLG and retention 4) Marketing engineer: builds AI agents that find your customers, write your outbound, test your ads, and get smarter every week Most companies already own the pieces in a stack of SaaS tools: - Dashboards - Call recordings - Content calendars - CRM But the learning is scattered: - Sales hears one version of the market. - Support hears a different one. - Product sees the usage. - Marketing sees the clicks. - The founder remembers the one call that hit him that week. Everybody then walks into the growth meeting with a different version of reality. One person pulls those signals into one system and turns them into pipeline. Hire that person, or become that person.
The Startup Ideas Podcast (SIP) 🧃42,251 görüntüleme • 1 ay önce
Struggling with Claude Code? Here's what Boris Cherny, the creator of Claude Code, recommends almost every time: 1) Use Opus 4.5 with thinking - It's smarter, uses fewer tokens, often ends up cheaper than smaller models. 2) Invest in your ClaudeMD - Just a text file. No special format. - Add every mistake Claude makes so it doesn't repeat them. 3) Give Claude a way to verify its output - Run tests. - Start a server. - Let it see the browser. A painter wearing a blindfolded won't paint well. Neither will an LLM that can't check its own work. Once the plan is good, the code is good.
The Startup Ideas Podcast (SIP) 🧃160,458 görüntüleme • 9 ay önce
One of the best skills to install right now is my friend Peter Yang's no AI slop skill. It's an editor. It hunts for the patterns that make writing feel AI generated and strips them out, while trying to preserve your actual voice. The second part is the hard one. Most writing tools make you cleaner and sand off the interesting parts, so everyone ends up sounding the same. You already know the smell. The grammar is fine, the syntax is fine, and it still reads like a keynote from a fake SaaS conference. It writes "it's not x but it's y." It uses "quietly" a lot. Here's how I run it: 1) Install it: npx skills add, then the GitHub link. 2) Write a rough draft yourself. An outline is fine, messy is fine. 3) Get your real points down, the ones only you would make 4) Ask the skill to remove the AI patterns and keep your voice. Step 4 only works if step 2 is real. If you ask AI to write the whole thing, there's no voice left to preserve. If you're building products, you're writing constantly. Tweets, landing pages, cold emails, launch posts, product updates, onboarding copy, investor updates. Nobody replies to say "this was written by AI." They just trust you less and keep scrolling. Write the messy draft, run the skill, then post it.
The Startup Ideas Podcast (SIP) 🧃40,507 görüntüleme • 1 ay önce
The best startup idea right now isn't software. It's a job. Old SaaS said: "here's a tool your team can use." Agent SaaS says: "here's a job your team never has to do again." That's the whole shift. A restaurant doesn't want dashboard software. They want the phone answered, reservations booked, and VIPs routed while the host seats tables. A plumber doesn't want a CRM. He wants every missed call picked up at 2am and turned into a booked job. Same demand. More revenue. Zero new hires. So the only question that matters: Can I do one annoying job better than a junior, faster than an agency, and cheaper than a new hire? Build that. The product is the job.
The Startup Ideas Podcast (SIP) 🧃74,149 görüntüleme • 4 ay önce
This is how we streamline AI implementation across 26 companies in a PE portfolio. Each company has 10 departments, and each department has 10 workflows. If you do them one at a time, you need "50,000 forward deployed engineers across three years." Our method is to group the companies by system of record. For finance, the 5 companies on NetSuite are one group. The 4 companies on Dynamics are a different group. Then you find what NetSuite can do and what it cannot do. You build only for the gaps. In one group, the processes are still different. The 5 NetSuite companies used 12, 9, 15, 18 and 13 steps for their finance process. The target is 6 steps, with agents at specific points. You will not often make all 5 exactly equal. You can come close if you map each process with the stakeholders before you change it. A single company or 26, sort the work by its system before you build.
The Startup Ideas Podcast (SIP) 🧃17,379 görüntüleme • 24 gün önce
Steal this startup idea Facebook Marketplace App Studio Facebook has 3B users. 16% use Marketplace monthly. That's 480M monthly users. 51% of ALL social media purchases happen there. People make their entire living flipping on Marketplace. Meanwhile: * eBay has thousands of apps (some acquired for $100M+) * Shopify devs earned $1B through their app ecosystem Yet there are ZERO real tools for Facebook Marketplace The API allows it. I vibe-coded one myself. It worked. Here's what sellers would pay for: 1. Price scanners (is this dresser underpriced?) 2. Arbitrage alerts (worth more on eBay?) 3. Auto-responders for leads 4. Listing optimization tools Productize what people are doing "in the shadows." Massive platform. No competition. That's the alpha.
The Startup Ideas Podcast (SIP) 🧃130,952 görüntüleme • 9 ay önce
One of the biggest arbitrage opportunities in AI is to build a "software factory." Here is the example: - I run an AI First Index with my Fortune 500 clients. - We interview their executives and score them on 16 dimensions. - Then we made a public version of it. Option one: open Claude Code, or Codex Say "build out the product." Spend hours refining and iterating until you finally release it Option two: go one level up and "build the factory for the thing." Ours has 5 parts: 1) Primitives. - Login. - Payments. - Social sharing. - You build each one only one time. 2) Promotion. - The owned media that sells the product is part of the machine 3) Memory - We map the roadblocks we had in a shared memory bank so never make that mistake again. - For example, we had a web hook problem. - "I want to learn through the mess." 4) Humans. - These loops are not autonomous. They are "very heavy-handed with humans." 5) Speed. - The first product is already profitable. - The next product is faster, better and stronger. The catch: This process is a "measure twice cut once", so make sure you take the time to build it right the first time Then use the same factory for your content engine, or for your new leads.
The Startup Ideas Podcast (SIP) 🧃40,176 görüntüleme • 2 ay önce
AI agents should add to the systems your client already uses. One client spent several years and $10 million on one ERP migration. Offer that client a move to an AI-native CRM, and "you've lost them." So we build our agents inside the systems of record. They read the data in Salesforce and do the work in Salesforce. When a human must be in the loop, the agent sends a message in Slack. The staff need no new training, so they use the agents faster. SMB clients with no CRM are the exception. You can put them on one. If you want to work as a forward deployed engineer, build on what your client already uses.
The Startup Ideas Podcast (SIP) 🧃15,996 görüntüleme • 25 gün önce
Buzz is Slack where the agents are teammates instead of integrations. Here's what that unlocks: Agents: - The ones you start with are just defaults - Edit them, add your own, write their instructions like a system prompt Swappable Harness: - This is the killer feature - Swap Claude Code, Codex, Goose under the same agent Context: - Every chat you've had with that agent carries over to the new harness - Global skills and skill files come too Think: Slack = the shape Agents = the teammates Harness = the engine you can change without losing anything "I get model fatigue" is a real thing. Buzz sits one layer above it.
The Startup Ideas Podcast (SIP) 🧃43,523 görüntüleme • 3 ay önce
One of the best markets to be building for is boomers. Mobile apps, desktop apps, agents, even hardware. There are 70 million-plus seniors in the US and almost nobody is building for them. AI is good at the exact things they need help with: hearing, mobility, memory, vision, and loneliness. Widen it down to 45 and it gets better. It won't be explosive growth. But if you want to build something that makes a couple of million, the people over 45 are the highest earners, they've built up the most savings, their kids have moved out, and they have the free time. Here's the setup: 1) Pick a category that only gets built for young people. Gaming, dating, fitness, anything with a cooler younger audience. 2) Find the pain point the older version has. Hearing, mobility, memory, vision, loneliness. 3) Build the older version of it. A 55-year-old isn't twitch-fast anymore and gets left out of anything competitive, so build the one you play for the fun of it. 4) Keep the gray hair and the balding off the landing page. They don't see themselves that way. They want a product that makes them happier and healthier. 5) Advertise on Facebook. People will tell you Facebook is dead. It's alive, just not in your age range, and a 55-year-old will hear about you and buy the same day. 6) Leave the targeting off and watch who buys. This gets better in 10 years. Gaming is getting more expensive and that whole audience is aging, so the people who can afford to play will be the older ones. Dads of Marathon exists. Nobody has built Grandpas of Marathon. Pick the category, build the older version, then let the ads tell you who's buying
The Startup Ideas Podcast (SIP) 🧃26,418 görüntüleme • 1 ay önce
