L3 Support & Incident Lead (Manila)

Job description

# Incident Commander ### Night shift · On-site, BGC Taguig · Individual contributor --- ## We're looking for an ER but for SaaS A patient comes in. The doctor examines them personally — checks what needs checking, forms a view from the evidence in front of them, right there, without a full workup. Then they command the room. *You run this. You administer that. We stabilise this before anyone goes to a ward.* They don't perform the surgery. But they decide what happens, and they decide it because they're the one who looked at the patient. That's this job. --- ## The 2am problem EviSmart automates case intake for over 2,000 dental labs across North America, Canada and Europe. A dentist sends a prescription, our platform reads it and puts it into the lab's system, and their work is waiting when they open up. We're step one in our customers' day. If a case isn't booked correctly within 15 minutes, the lab has nothing to work on. Their business hours are our overnight. And overnight, our engineering and dev teams are asleep. So when a lab in Chicago opens their system and their cases aren't there, you are the entire response. You look first. Then you decide. Then you tell people what to do. --- ## What you'll actually do **Diagnose it yourself.** Open the application logs. Query the database. Check what shipped in the last release. Work out what's wrong from evidence, not from a runbook. **Then command the response.** Direct Application Support, DevOps and Customer Support on what to do next — your call, based on your findings. **Ship the workaround.** Most overnight problems need a permanent fix only the dev team can make, and they're not there. So you build the stopgap that keeps the customer working until morning. This is the core of the job, not an occasional ask. **Hand off clean.** The day team should never have to reconstruct what happened overnight. **Train your backup.** One Application Support specialist, taught to run real night triage on their own. **Stop the repeats.** The same problem three nights running is a platform problem, and it's yours to push to a permanent fix. --- ## What we'll ask you to prove We'll ask for a specific example of each of these. - You've personally diagnosed a production issue by reading application logs, querying a database, comparing releases, or tracing an API call — not by convening a call and asking what other people see. - You've built a workaround yourself that kept customers working while someone else built the permanent fix. - You've directed engineers who don't report to you, based on findings you produced. - You've owned a decision alone, under pressure, with nobody available to check with. - You can teach a peer to do what you do without formal authority over them. - You can give a customer-facing team an accurate, plain-language status update while an incident is still live. Deep DevOps expertise isn't required — that's Jester's team. You need enough to apply a safe first fix and know exactly when to hand off. **You're probably in** L2 or L3 application support, production support, technical escalation, or platform support. What matters is that you've been the one looking at the system, not the one coordinating the people who look at it. --- ## Three things to know before you apply **It's night shift.** Roughly 11pm to 8am. That's the whole reason the role exists. **It's 100% on-site**, at our BGC office in Taguig. **It's an individual contributor role.** No direct reports. The command in this job comes from being the person who knows what's wrong — not from an org chart. If any of those is a no, this isn't the right role, and we'd rather you know now. --- ## How we'll set you up You report to Jester Tolosa, our FPE for DevOps and Cloud. The first 90 days are a structured ramp. You'll spend real time learning the platform and the team before you're expected to run night incidents alone, then build to full independent ownership — including a trained backup — by the end of it. Our documentation is thin. Part of the job is writing down what doesn't exist yet. If you want a mature environment with a documented answer for every scenario, this isn't it.