About RecoverStack

I'm Yarin Goldstein. 24, based in Israel, and RecoverStack is my first real business.

The 5-year setup

I got my first computer when I was 5 and it's basically been a match since then. Making slides, photography, video, eventually code. In tenth grade (2018) I started learning C# at school, quickly ran ahead of the class, and got curious about how the big players (Google, YouTube, Facebook) actually built their products at scale. School pace wasn't going to get me there, so I went looking on my own.

For the past 5 years I've been programming professionally. NDAs keep the specifics vague. What I can say is I got a lot of responsibility young, spent a long time on high-traffic systems with tens of thousands of live users, and learned how dev teams should actually function under real pressure.

The best moments inside that job were the sub-products. Coding something from zero, talking 1:1 with the customers using it, getting unfiltered feedback. That loop is what I'm trying to rebuild with RecoverStack, at a scale I own.

Sitting at the bottom of the pyramid, or playing middle manager, both had ceilings. After 5 years I just wanted to do my own thing.

Hevreman, and meeting Claude

My first real attempt was Hevreman. A smart carpooling app that matches drivers and passengers heading to the same destination (events, work, nights out). I don't own a car, and my friend group constantly has the "who's taking who home" problem. I figured an app could handle the matching.

I got the route optimization done, the server done, the front-end almost done. Then I started integrating map services and SMS verification and watched the spend ramp up. I wasn't scared of that part. I had a dedicated business and investment budget that was years in the making for exactly this moment. What I realized was something else: I'd been thinking at scale technically, but not economically. I had no idea how much ad revenue could actually cover, or whether it could cover anything at all.

Around that time I was skeptical of AI coding tools. A friend of mine was paying $100/mo for Claude Max and I thought he was crazy. Eventually I tried it. The experience was heaven and earth compared to what I'd been using. It's a bit like moving to the back seat of your own car and letting someone else drive. Scary the first few minutes, much easier the rest of the ride.

At some point I thought: maybe I should challenge the model on something other than code.

The verdict and the pivot

I asked Claude to run a profitability assessment on Hevreman based on everything it knew about the product and my plan to monetize with ads. Five minutes later it came back with a detailed report. I read the whole thing and realized Claude wasn't just going along with my assumptions. It was opinionated, it pushed back (ad revenue won't work on a web product the way you're describing), and it could back its claims.

Then it offered alternatives.

I decided to pivot instead of grind through a business-model shift. Changing the model on Hevreman would mean reaching for a different audience, and that wasn't what I wanted to spend the next 9 months doing.

So I asked the next question: what should I build?

How RecoverStack happened

I gave Claude my real situation. Solo. No team. Running this in parallel with a day job. Real investment budget. And I wanted to set the parameters deliberately extreme: one week to ship, $5K in revenue within a month of launch. The logic was that if Claude came back with ideas that survived those constraints, the worst case was still a very good idea.

One of the answers was a dunning tool for bootstrapped SaaS on Stripe.

Before that conversation, I didn't know this problem existed. I assumed Stripe handled all of it. Finding out they don't, and that the average SaaS loses roughly 5 to 10% of MRR to failed payments silently every month, was genuinely shocking. If I didn't know about this leak, other founders don't either. Churn Buster and Churnkey, the two main tools in the category, both start at $249 to $250/mo. That leaves everyone under $100K MRR without a real option.

RecoverStack is the tool I'm building for that gap. Flat $49 to $149/mo. Decline-code-aware retries. Transparent public pricing. Public metrics.

Where I am right now

I'm the lead on this. Claude is the force multiplier. I call the pattern we use the Orchestrator: I write strategy and briefs, Claude writes most of the code, I review and integrate. A JWKS migration I'd have expected to cost me a full working day took 15 minutes of my time while I was making lunch. It feels like I moved from worker to manager overnight, and I don't want to go back.

I'm not a veteran operator. I've never run a business before. I'm figuring out marketing, pricing, and customer conversations in public, in real time. What I do have is 5 years of serious dev experience, a real investment budget, Claude as a thinking partner, and stubbornness. The last one isn't on a résumé. It's the one that matters here.

Founding cohort is free through August 1, 2026. The first 30 seats lock in 40% off for life after that.

If you want to watch the numbers, they're live at recoverstack.dev/metrics. If you want in, the waitlist is at recoverstack.dev/early-access. If you just want to trade notes, I'm on LinkedIn at linkedin.com/in/yarin-goldstein-9295783a3.

Yarin