Yevgeniy Melnik

Every profitable service business I’ve looked at shares the same pattern: something that’s mandatory, unpleasant, and complex enough that people will pay to outsource it. Taxes, compliance, vehicle inspections, estate planning — the formula is always the same. If it’s required by law, hated by the person responsible, and easy to get wrong, someone is already making money doing it for them. Whenever I hear about a regulation getting more complex, I don’t see a burden. I see the next business.

I pulled my Amazon order history to check my coffee buying pattern. Seventeen bags over 165 days. One bag every eleven days, but I was buying one-off instead of subscribing. Subscribe & Save would have saved me $82 a year on coffee alone — and Amazon never told me, because their incentive is for me to forget and pay full price. Businesses have sophisticated inventory and reorder systems. Consumers are flying blind on their own household supply chain. Someone should build the personal procurement layer — your purchase history, your consumption patterns, your reorder timing, your cross-retailer view. Not what the retailer wants you to buy. What you actually need and when.

My kids’ school photos cost $50 per package and the quality is terrible. The photographer shows up once, herds 400 kids through a backdrop in 3 hours, and the parents get stiff poses with bad lighting. The obvious startup move is to build a better platform — AI sorting, digital delivery, face matching. But that’s the wrong problem. The real problem is operational: the photographer is incentivized to maximize throughput, not quality. Fix the ops — better lighting rigs, more time per kid, a second pass for retakes — and you win without building a platform. The technology already exists. The execution is what’s broken.

Every time someone moves, they spend weeks tracking down every service, subscription, bank, insurance company, and government agency that has their old address. There’s no master list. Every person rebuilds the same checklist from scratch and still misses half of it. Your email already knows every company you’ve ever interacted with — the confirmation emails, the shipping notifications, the account alerts. A tool that scans your email headers, identifies every entity that has your address, and generates your personalized “who to notify” list would turn a month-long anxiety project into a ten-minute task. The fact that nobody has built this well says more about the difficulty of email parsing than about the value of the solution.

In every regulated industry, complexity creates information asymmetry. Whoever bridges that gap takes a cut of enormous transaction volumes. Medical bill negotiation: hospitals charge opaque prices, patients don’t know how to push back, a negotiator saves 50-80% and takes a percentage. Property tax appeals: data shows the assessed value is wrong, you file on contingency, win rate is over 60%. Government IT staffing: match cleared professionals to agencies, mark up 40-60%. The pattern is always the same — mandatory spend, confused buyer, knowledgeable intermediary. Software makes the intermediary faster, but the judgment call is what the client pays for. These aren’t sexy. They’re durable.

Job boards are designed for companies that pay to post listings. Most local businesses — the restaurant, the auto shop, the warehouse down the street — don’t post anywhere. They put a sign in the window. If you’re looking for work in your neighborhood, there’s no way to see what’s available without physically driving around. A tool that maps local businesses and shows which ones are hiring would bypass the entire application-tracking-system problem. The people who need local work the most — hourly workers, parents with commute constraints, people re-entering the workforce — are the ones most underserved by the current job search infrastructure.