“Our checkout latency spiked every lunch rush. Web Maple Core traced it to a missing composite index and an ORM prefetch that loaded an entire price-history table. After the index landed, p95 dropped enough that we cancelled the emergency shard plan — though we still owe them a cleanup of two leftover debug views.”
Hana Wu
Platform engineer · Taipei e-commerce · Query Performance Review
“They refused to ‘just rename tables overnight.’ The expand–contract plan took longer on paper than our previous contractor’s promise, but we never had a Friday-night rollback. Communication in English and Mandarin on the cutover bridge helped our ops crew.”
David Lin
CTO · regional payments startup · Database Refactoring
“The schema health report was blunt about our nullable foreign keys. A few findings felt pedantic at first; two months later one of those gaps would have broken an acquisition data room. I wish we had budgeted an extra day for the briefing.”
Priya Natarajan
VP Engineering · logistics SaaS with TW HQ · Schema Health Assessment
“Pairing on the migration scripts taught our mid-level developers more than a training course would have. The only friction: staging data was thinner than production, so one plan estimate was optimistic until we refreshed the dump.”
Kenji Sato
Tech lead · cross-border retail · Database Refactoring
Extended story: nightly warehouse load in Kaohsiung
A logistics client ran a four-hour ETL that sometimes spilled into the morning dispatch window. Web Maple Core reviewed the extract queries against their operational MySQL cluster, identified a correlated subquery rewritten as a join, and sequenced index builds during a Saturday maintenance slot.
Outcome after two consulting weeks: the load finished inside the warehouse window on fifteen consecutive nights. The team kept a short runbook we co-authored for the next schema change. Residual risk remains on a legacy reporting database they plan to retire next quarter — noted explicitly in the handoff, not hidden.