He Fired Almost Everyone and Raised a $30M Series A

Rafael Broshi launched a crypto insurance product in 47 states, backed by a real carrier. Everyone called it genius. Nobody bought it. He shut it down, rebuilt with two engineers, and Notch just raised a $30M Series A — 2x the current median.

Key Facts

• Series A raised: $30M from Headline — 2x the current median Series A (Pablo Srugo)
• Burn at rebuild stage: $10K–$20K/month with 2 engineers (Rafael Broshi)
• PMF signal: Won a 7-figure enterprise POC entered mid-race with less than one week of prep (Rafael Broshi)
• Pivot trigger: $2M left in the bank after shutting down crypto insurance product (Rafael Broshi)
• Warning sign: 'That idea is genius' — great feedback that usually means no real pain exists (Rafael Broshi)

The PMF Moment: Winning an Enterprise POC From Behind

Notch's PMF moment came when they entered a large insurance carrier POC mid-race — without full context or preparation time — and became the front-runner after their second iteration. Winning without rebuilding the product confirmed that what they had already fit the market.

Most founders look for product-market fit in customer surveys or NPS scores. For Rafael Broshi, it arrived during a high-stakes POC with a major insurance carrier — one Notch entered late, through the back door, with less than a week to prepare.

The POC required three product iterations. After the second version shipped — built over a sleepless weekend by six or seven people — the client's tone shifted entirely. Notch went from 'we'll give them a shot' to front-runner status in days.

That shift was the signal. As Rafael put it: if you can enter a competitive enterprise evaluation late and still lead the race without rebuilding your product, you have something real. The first seven-figure contract followed, and it confirmed Notch's architecture — built to handle complexity without degrading quality — was genuinely differentiated.

"If we get to an enterprise kind of POC and we figure out a way to be a front runner in the race with less than a week of preparation, that means that we have a good product that already fits those kinds of clients." — Rafael Broshi

"After we brought in the first version, suddenly the entire conversation kind of changed from us being the underdog and giving us a chance just because they wanted another vendor in the POC." — Rafael Broshi

One Enterprise Deal Doesn't Mean Ten — Here's How to Tell the Difference

A single enterprise win signals PMF only if you won without heavy customization. If you rebuilt your product to fit that one client, you may have proven your team's execution speed — not that your product fits the market. The test is whether the client's requirements were generic or unique.

Closing an enterprise contract feels like the hard part is over. It's not. Rafael is direct about the trap: many startups get one or two enterprise clients and become their personal dev shop, customizing everything the client asks for, fast. That's not product-market fit — that's outsourced development.

The green flag is the opposite. When the things clients ask for are generic — fast responses, large corpus support, audit trails, on-prem deployment — those are table-stakes requirements any serious buyer in the space will have. Winning on those terms means the next ten deals look the same.

Rafael frames it simply: 'If at the end of the day you're saying, hey, I did not have to change anything — I just had to prepare the demo in the right way, or I just had to do two integrations to meet their requirements — then that's totally different.' That's the distinction between a repeatable product and a professional services engagement dressed up as a SaaS deal.

"Every enterprise that decides to work with a small startup wants speed and customization. If you won the POC because you customized everything about your system to fit what the client wanted — that should be a yellow light in your head." — Rafael Broshi

• Yellow flag: you won because you customized everything to fit one client's needs.
• Green flag: client requirements were generic — the same things every enterprise asks for.
• True PMF means the next buyer wants the same thing, not a new version.
• Winning an enterprise contract is the easiest part of working with one.

'That Idea Is Genius' Is a Warning Sign, Not a Compliment

When everyone calls your idea genius, that usually means nobody actually needs it. The feedback you want from potential customers is skepticism — 'isn't everyone doing this already?' That means the problem is real, budgeted, and worth solving right now.

Notch's first product — crypto insurance for digital assets, launched across 47 states with a real carrier — checked every external box. The carrier was legit. The regulatory work was real. And everyone who heard the pitch said it was genius. That was the problem.

Rafael's conclusion: 'If you were in a coffee shop and you heard someone talking about an idea that they're actually executing on, a startup three months ahead of you — would you be happy or completely bummed? Of course you'd be bummed. So why don't you want to tell people about your idea? Put it everywhere. No one's going to want to do it.' The fear of competition is usually inversely correlated with whether a real market exists.

When Notch pivoted to AI for customer experience, the feedback flipped entirely. Nobody said it was genius. They said it seemed crowded. That skepticism — 'isn't everyone already doing this?' — turned out to be exactly the right signal. The pain was real, the budget existed, and the market was opening.

"The product was great. I mean, everyone said it was great. It sounded genius. The only problem was — and that's a big problem — there was no product market fit. There was no real pain that we were solving." — Rafael Broshi

"Right now, the main feedback we got is 'isn't everyone doing it?' And that's a bummer, but you shouldn't care. Because innovating usually means innovating in one place." — Rafael Broshi

How to Pick the Right Wave: Crypto Faded, AI Didn't

Ride waves where budget already exists or will exist in 6–12 months. Build too early and you burn out before the market arrives. The right signal is an 'earthquake' — a visible market shift where incumbents' existing investments are suddenly obsolete and the playing field resets.

Notch's pivot from crypto insurance to AI customer experience wasn't obvious in the moment. Rafael is clear on that: 'Everything looks obvious in hindsight. When a wave starts, it's incredibly hard to understand.' NFTs looked like a wave. ChatGPT also looked like it could be hype. The difference was the speed and depth of enterprise adoption.

His framework for evaluating waves: look for an 'earthquake moment' where existing technology investments are wiped out and everyone starts from zero. When ChatGPT launched, years of NLP investment by incumbents became irrelevant overnight. That reset created an opening. 'You had the exact same shot of succeeding as Salesforce or Zendesk,' Rafael notes.

The key constraint is timing. Building for a wave that arrives three years late means someone who started two years after you will have better technology and less technical debt. You need to be early enough to build, but not so early that the market never shows up while you're funded.

"You want to look at a market that is opening. It's like an earthquake — when you see the earth shatters and you start seeing a huge gap. That's exactly what happens now in AI for customer experience." — Rafael Broshi

"Even if you don't run out of money by the time the wave starts, someone that started building two years after you will have better technology and less tech debt." — Rafael Broshi

Getting Enterprise to Take Your Call: The Tactical Playbook

There's no formula. Use every relationship you have — investors, employees, family — and ask sincerely for help. Work both top-down and bottom-up simultaneously. And go to your dream customers first, not after you've proven yourself with smaller logos.

The most common mistake Rafael sees: founders building for six months before talking to the company they actually want to sell to. His advice is blunt — go to your dream customers first. If they're not interested, you know now. Waiting until you have traction to approach them just delays the most important feedback you'll get.

For getting in the room, hustle through every connection available. 'Use all of your investors, use all of your employees, use your father, your mother, your brothers, your sisters. Everyone wants to help. I've never had anyone say no to me.' The key is sincerity — asking directly, 'I really need your help closing a sale here,' works better than positioning yourself as superior.

On fake POCs: if the person running the POC doesn't own the budget and there's no defined success criteria, it's not a real POC. Rafael's test is simple — ask what happens if you succeed. If there's no clear next step tied to a budget, you're doing unpaid development work. Charge for POCs, meet the business owner, and define KPIs upfront.

"Who are your dream customers? Go talk to them. Well, we don't have a product. No — that's the wrong way to go. Don't go the long route. Talk to your dream customers first." — Rafael Broshi

"If people want your product and you send an offer, they're not going to move out just because you wrote a number they don't like. Companies understand that if they want a good partner, they need to pay." — Rafael Broshi

• Ask directly for introductions — sincerity outperforms positioning.
• Work top-down and bottom-up in an enterprise account simultaneously.
• Charge for POCs and define success criteria before starting.
• Meet the business owner — tech evaluators alone cannot buy.
• Go to dream customers first, not after earning smaller logos.

Real PMF vs. Strong Execution — How to Tell the Difference