Rob Bailey retweeted
yesterday, I met a YC founder who’d just pivoted.
the kind of founder you could see raising a $15M series A in the next 6 months.
asked him how he was finding customers for the new product.
his answer:
“i look for people who just inherited the problem.”
thought he meant newly promoted executives.
sort of...
he finds people who recently started a role.
then looks for evidence that they’re trying to fix something his product solves.
a new VP of sales posting about rebuilding outbound.
a new head of ops asking how other teams handle onboarding.
a new marketing lead discussing attribution tools.
the job change gets them on his radar.
what they say next gives him a reason to reach out.
because someone starting a new role is usually doing a few things:
> asking questions.
> reviewing what’s already there.
> deciding what needs to change.
and sometimes, they’re doing all of that publicly on linkedIn.
imagine this comment on linkedIn:
“3 weeks into the new role. great team, but our reps spend way too much time updating spreadsheets. fixing that is our top priority.”
most people see a career update.
he sees a specific problem, someone who owns it, and a reason to start a conversation.
his message would be:
“hey [name], saw your post about the spreadsheet overhead. where’s most of the manual work happening? prospecting, CRM updates, or reporting?”
that’s it.
no calendar link.
no unsolicited demo.
one question they can answer in 10 seconds.
if they say “reporting,” and his product only fixes prospecting, he moves on.
if they say “prospecting, we’re doing everything manually,” there’s a conversation worth having.
here’s the playbook:
step 1: find people who recently joined roles that own the problem you solve.
step 2: check the company fits. right team size, relevant use case, realistic need.
step 3: read their recent posts and comments. look for a priority, a question, or a frustration.
step 4: reference the specific thing they said.
step 5: ask one useful question before asking for a meeting.
you can use @GojiBerryAI to help surface prospects through job changes and topic engagement, then personalize outreach around those signals.
now it’s your turn.
find someone who just inherited a problem you can solve.
start the conversation.
go close some deals.
Rob Bailey retweeted
99% of salespeople ask "random acts of questions."
They come off as a checklist.
They don't GUIDE the buyer through a sequence of 5 realizations.
That's the key.
And that's what star salespeople do in the modern era of selling.
Here's how it works.
Rob Bailey retweeted
I’m the CEO of Product Hunt and see 100s of launches a week. My current thoughts on new infra begging for startups:
1) Voice AI - infra for voice agents is getting really good w @pipecat_ai @Vapi_AI @DeepgramAI @ElevenLabs et al. and yet I rarely see good voice tech in the real world.
2) Decision Models - Jev / @typesafeai leading the charge. It reminds me of the mech-building in Zelda TOTK - an explosion of possibilities!
3) Personal Agents - they work now. And they need tools, apps, MCPs.
4) Extensions (OpenAI). We all live in agentic interfaces now. It’s like a new platform. Someone needs to build a Slack replacement that lives in Codex, for example.
Dear Friends: We are hiring! We will be ramping up fast serving our customers in the coming months. Hiring Developers (Python+AI Dev Tools), FDEs and AEs. We are moving extremely fast. DM me. NO RECRUITERS. Also, no anonymous x accounts. Must be US time zone or BST.
Dear NYC Friends: If you care about law and order in NYC, I invite you to learn more about Saritha.
New poll out and James is only up 6:
🔵 Letitia James 51%
🔴 Saritha Komatireddy 45%
That’s basically half the lead she had last month.
People who actually watched the debate went hard for Saritha. Independents are moving her way too.
James has used the AG’s office for politics while crime and fraud keep hitting regular people. Saritha’s a former federal prosecutor who’ll actually do the job.
@sarithaforny has my support. Vote November 3. 🇺🇸
Getting ready for launch! Ramping up with Josh Whitfield from Contentmarkettingmedia! (He was on stage at the Clay conference.) CC @@vxanand
UPDATE: Please include in your DMs
- last 3 cos you worked at / founded
- what are your biggest wins?
- what have you built in AI/Agent that is in production
- side projects, consumer apps, SaaS = not relevant
- MUST be americas NOT India or ME (sorry)
If you are applying for a role at our company, you can't have an anonymous X account. No exceptions.
A new member of our team is in week 3. A report he produced for a lead customer radically changed/evolved our product roadmap already. ♥️
Zero to one: Hire amazing people. Work with smart lead customers. Ship what the customers want extremely fast. Give the amazing people whatever they need to get stuff done.
Rob Bailey retweeted
Before you launch anything, have Opus 5.5 send 5 subagents in as your first users.
Paste this into Claude Code:
"Act as my launch QA lead. Spin up 5 subagents, each a different first-time user:
1. The impatient one: skims, clicks the first button, quits at the first error
2. The skeptic: reads pricing, terms and privacy before signing up
3. The mobile user: does everything on a 375px screen on slow 3G
4. The non-native English speaker: only follows what the UI literally says
5. The breaker: pastes junk into forms, hits back mid-flow, double-clicks everything
Each one opens the app in a real browser, goes from landing page to first success, and screenshots every moment they get confused, stuck or annoyed.
Merge everything into one list ranked by how many users hit it. Fix it all in one pass, then rerun the 5 users to prove it's fixed."
Your first 100 real users won't tell you why they left. these 5 will.
(Works best with a browser MCP connected, Playwright or Chrome DevTools)
Rob Bailey retweeted
My cofounder is a beast.
He built the setup that helped us scale to multiple millions in ARR.
And the system behind it is surprisingly simple.
Here’s his playbook 👇
1/ Start with 3 LinkedIn profiles.
His, mine, and a team member’s.
All creating content around the problem our product solves.
2/ Get 3 screens.
A laptop, a monitor, an iPad.
Nothing fancy. You can see how that escalated 😂
3/ Create lead magnets your ideal customers actually want.
Templates. Playbooks. Checklists.
His target: 2 lead magnet posts per profile, per day.
Then reply to comments and keep the conversations going.
4/ Ask Claude to build a chrome extension to help respond to comments faster.
Less time on repetitive work. More time talking to potential customers.
5/ Reinvest the first revenue into distribution.
Pay creators whose audiences match your ICP to share your lead magnets.
Track what brings qualified leads. Double down on what works.
6/ Use @gojiberryai to find relevant leads, filter for your ICP, and run multichannel outreach to book demos.
7/ Keep both acquisition channels running.
LinkedIn content brings people in.
Outbound gets you in front of prospects who haven’t found you yet.
The initial setup can cost less than $1,000 if you do the work yourself.
8/ Reinvest revenue from closed deals into more LinkedIn distribution.
9/ Repeat.
That’s the system behind his desk.
It is becoming f*cking dangerous
A 23-year-old dev from China just ran 300 Claude Code agents in parallel and hooked them to a verification layer so none of them could bullshit him
He opens the live dashboard:
- 300 agents writing and checking at the same time
- Every output verified against the source in ~100 ms
- Three passes over 100 company profiles → zero errors left
- Everything dumped straight into an Obsidian vault
The trick is the split: one model does the work, a cheaper one decides if the claim is actually true. The loop doesn’t stop until everything checks out
An analyst team would bill weeks for this. He got it in one run
Same kind of Second Brain + swarm setup that keeps circulating. Bookmark the full breakdown
Rob Bailey retweeted
We're building an AI software factory for clients at @LimestoneHQ.
The goal: take a ticket from requirements to a tested pull request using AI agents, with human approval and governance built into the process.
Here's what goes into the factory:
1. Orchestration — Coordinates agents, triggers workflows from tickets, and manages execution.
2. Spec-Driven Development — Turns requirements into specifications, technical plans, and testable tasks before code is written.
3. Agent Runtimes — LangGraph.js + Claude Agent SDK for executing specialized agents.
4. Execution Infrastructure — AWS Fargate with pre-warmed, isolated sandboxes for parallel coding tasks.
5. LLM Gateway — LiteLLM / AWS Bedrock for model routing, API keys, rate limits, and token budgets.
6. Governance — Anthrive for centralized policies, prompt enrichment, and guardrails.
7. PR Review & Grounding — AI code review surrounded by deterministic validation. Every finding needs real evidence.
8. Code & Knowledge Graphs — Maps code dependencies, architecture, domain knowledge, and decisions so agents retrieve relevant context.
9. Memory — Postgres for execution history, checkpoints, and feedback from previous runs.
10. Observability — Langfuse + ClickHouse for tracing model calls, latency, tokens, and costs.
11. Testing & Evaluations — Automated tests, golden-PR regression suites, and versioned prompt evaluations.
12. Developer Intelligence — Measures engineering cycle time, AI usage, cost, and effort at the feature level.
13. Integrations — Jira, Azure DevOps, GitHub, ClickUp, and existing CI/CD pipelines.
The workflow:
Ticket → Spec → Plan → Build → Test → Review → PR → Human approval → Ship.
The important part isn't getting an AI agent to write code.
It's building the infrastructure around agents that makes software delivery repeatable, governed, and measurable.
DM me if you want your own self-hosted factory built for your enterprise.

















