Jordan Ellis, thirty-one, wrote production Java and Go services serving millions of requests daily at a company whose name appears on every “best places to work” list. He also attended four standing meetings weekly that could have been status posts, on-called one weekend per month, and spent promotion cycles documenting impact for committees that met quarterly. In March 2025 he resigned with eight weeks savings and no plan beyond “not this.”
We interviewed Jordan about why engineers leave big tech when compensation peaks, what surprised him about smaller companies, and whether he regrets trading RSU refreshers for sanity. He spoke candidly; his current employer knows he gave this interview and did not review quotes.
Jordan’s partner supported the quit but required a written budget before he handed notice. That spreadsheet—rent, runway, health insurance COBRA—became the most important engineering document he wrote that year.
We also asked Jordan whether he would return to big tech. His answer was conditional: yes for a specific problem and team he believed in, no for vesting alone, and definitely not without six months runway banked first regardless of headline compensation.
Golden Handcuffs and the Slow Leak
Jordan’s total compensation exceeded three hundred thousand dollars annually at peak—base, bonus, vesting stock. Leaving meant walking away from unvested grants he had counted on for a down payment timeline. He stayed two years longer than his gut recommended because the math felt irresponsible to ignore.
What changed was not one bad manager or failed launch. It was cumulative friction: internal tool migrations that existed for compliance theater, reorgs that reset team context every eighteen months, and the sense that his best work disappeared into infrastructure consumers never saw. He felt like a component.
Stock vesting schedules created perverse retention: each quarter he calculated walking-away cost and felt golden handcuffs tighten. He uses the phrase without irony now when mentoring engineers who think misery is mandatory at certain logos.
“I stayed two extra years for RSUs my gut had already quit on.”
The Exit Without a Net
Jordan did not have another big tech offer—he stopped interviewing there entirely. He took three months off, which terrified him financially despite years of high income with lifestyle inflation to match. He fixed his bike, contributed to open source without OKR alignment, and discovered how much cognitive space on-call rotation had consumed.
Recruiters from startups found him on LinkedIn. He was skeptical of “wear many hats” language until he interviewed at a fintech rebuilding payment reconciliation and met engineers who owned features end to end—including customer-visible outcomes.
Unemployment was never an option he considered dignified—until three months off rewired his definition of productivity. He reads more fiction now. He deploys less on weekends. He calls both improvements.
“Three months off showed me how much on-call had eaten my brain.”
Life at a Forty-Person Company
Jordan joined as a senior backend engineer at one hundred forty-five thousand dollars base plus modest equity—on paper a massive pay cut. After-tax cash flow dropped; stress dropped faster. He ships code that reaches production in days, not quarters. He joins customer calls when bugs affect merchants. He knows every person on the engineering team by coffee preference.
Tradeoffs are real: fewer dedicated SREs means he debugs production incidents at 2 a.m. sometimes. Benefits are fine, not lavish. No internal AI copilot suite—he uses the same tools indie developers use. He finds that oddly freeing.
Startup equity is lottery tickets he values honestly. His fintech grants might be worth something; might not. He traded known RSU decay for unknown upside and sleeps through the night either way.
“Paper pay cut. Faster shipping. I know every engineer by coffee preference.”
Skills That Transferred—and Skills He Relearned
System design instincts from hyperscale environments transfer: Jordan still thinks about failure domains, idempotency, and observability. What he relearned is product judgment—saying no to technically interesting work that does not move revenue or retention. Big tech had product managers absorbing that filter; startup engineers absorb it themselves or waste runway.
He mentors engineers considering similar jumps: savings matter more than Twitter bravado suggests. Six months runway minimum if you quit without a offer. Interview smaller companies for how they handle incidents and tech debt, not just Kubernetes trivia.
Incident pages still spike his adrenaline—but now he owns fixes end to end instead of routing tickets across three teams. Responsibility increased; bureaucratic heat decreased.
He rebuilt exercise habits post-exit—gym three mornings weekly—for the first time since college. Big tech had gym reimbursement he never used because on-call anxiety colonized mornings. Small detail; large signal about reclaimed time.
“At startups you relearn product judgment—or you burn runway on cool tech.”
Regret Check-In at Twelve Months
Jordan does not regret leaving. He regrets not leaving earlier and keeping more RSU cash unspent on a lease he took to commute to an campus he rarely needed. He may return to a larger company someday—for the right problem, not the right vesting schedule—but not from desperation.
Promotion title is Staff Engineer on LinkedIn instead of Senior SDE II at a famous logo. His parents finally stopped asking when he is going back to the big company. His sleep tracker improved more than his net worth last year.
Parents asked about prestige loss for eighteen months. Thanksgiving 2025 they stopped when he paid for dinner without checking his balance first on a smaller salary.
“Staff title at forty people beats Senior SDE II misery at a famous logo.”
What He Tells Engineers Still Inside
If you are miserable and well paid, you are not ungrateful—you are human. Big tech compensation is designed to make leaving feel financially irrational. Run the irrationality math anyway: health cost, years you cannot get back, whether your learning curve flattened.
You do not need a heroic quit story. Jordan’s was quiet, scary, and ordinary. That is still a valid path out.
He tells engineers still inside: you are not your vesting schedule. If therapy co-pays are line items you hide from coworkers, include them in your compensation math.
Jordan notes interview loops at startups varied wildly: one asked only leetcode; another asked how he would prioritize merchant-facing bugs versus internal refactors. He preferred the latter and joined that company.
COBRA cost twelve hundred monthly for three months—budget line he underestimated. He recommends engineers model healthcare explicitly when comparing big tech benefits to startup offers; sticker salary shock is only half the spreadsheet.
“Run the math on health cost, not just unvested stock.”
Jordan’s story fits a pattern hiring data confirms: senior engineers leaving marquee employers for impact, ownership, and calmer calendars—not always for maximum cash. Big tech remains a rational choice for many. So is leaving. The mistake is assuming compensation is the only variable that compounds.
Big tech remains a rational choice for many engineers Jordan respects. His exit is data point, not sermon. The variable he underestimated was cognitive bandwidth returned—not dollars lost.
Jordan still mentors two former coworkers quietly exploring exits. He shares his budget spreadsheet template—not his verdict. Some stay and feel better after honest conversations; some leave. Both outcomes beat silent misery funded by vesting schedules.
His fintech employer raised base ten percent at year one—a validation that impact and compensation can coexist outside marquee logos, though Jordan notes not every small company survives to grant raises.