What “senior” actually buys you.

“Senior engineer” is the most inflated title in software. On a CV it means anything from “ten years of payment infrastructure” to “was the oldest person at a startup.” When an agency is betting a client relationship on outside delivery, the difference is everything.

Here is what the word should mean — and what it looks like in a system you can go and inspect.

Anyone can build the flow where everything works. The craft is in the other paths. A membership system we run had a rare, expensive failure mode: payment succeeds, account creation fails, and a paying customer is left stranded — invisible until they complain.

The senior answer wasn’t “monitor it more.” It was to make the system repair itself: detect the orphaned payment, rebuild the membership, notify nobody because nothing is wrong anymore. Failure modes that fix themselves are the signature of engineers who have been paged at 2am and never want to be again.

Billing code is where junior optimism goes to die. Two payment rails that must reconcile into one customer record; renewals that must not double-charge; refunds that must never be casual. The engineering isn’t glamorous — it’s idempotency, reconciliation, audit trails, and a deep suspicion of edge cases.

When you’re evaluating a delivery partner, ask one question: “tell me about a billing bug you’ve prevented.” The answer tells you whether they’ve worked on systems where mistakes cost actual money — or just systems where mistakes cost a refresh.

Apple’s App Review is a brutal teacher. Categories with strict rules, rejections with one-line explanations, release dates that don’t move because a reviewer is having a slow week. Getting a subscription app through review in a hard category isn’t a coding skill — it’s scar tissue: knowing the rules, designing around them from day one, and handling the rejection cycle without panic.

The same is true of payment providers, of OTA channel integrations, of WhatsApp Business templates. Senior engineers have been through the gatekeepers before — which is why your timeline survives them.

A weekly demo, every week, even when it’s uncomfortable. Documentation written as the system is built, not confessed at handover. Conventional code that the next engineer reads without a decoder ring. Estimates given with reasons attached.

None of this photographs well. All of it is why some projects land on time and quietly, and others become the story an agency tells about why they don’t outsource anymore.

The good news: this is all inspectable. Ask for the live systems, not the slide deck. Ask what broke and what was done about it. Ask to sit in on one weekly demo. A safe pair of hands will show you; a title will change the subject.