Your AWS production is down or bleeding money. One senior engineer jumps in, finds the real cause in the infrastructure or the code, fixes it on prod, and hands you the written root-cause analysis. If the fire doesn't go out, you don't pay.
"AWS Support does not include custom code development, software debugging, system administration tasks, or remote access to customer systems and accounts."— AWS SUPPORT FAQ
When production burns, your engineers are stuck: the fire lives in a layer they don't own, or across two layers at once. Your cloud partner opens a ticket and "looks into it." Meanwhile more than half of significant outages now cost over $100,000 (Uptime Institute, 2024). The person who answers here is the person who fixes the system, and it starts the same day.
Tell us what is burning and what failure costs you. We agree what "fixed" means before the work starts — that is the measure for no-fix-no-fee.
A senior engineer gets hands on the system the same day, reads the telemetry, the architecture, and the code, and states the next step.
Written war-room updates on a fixed cadence. The price includes exactly two calls: a 30-minute kickoff and a 45-minute RCA readout. Everything between them is fixing, not talking.
The engagement ends when the system is stable and the written root-cause analysis is in your hands, closing with the list of what must change.
After the fire: the RCA closes with the list of what must change so it does not happen again. If you want us to execute it, that is separate work with its own quote: hardening, load testing before your next peak, a security or architecture review, cost work.
Declare, we respond, triage, fix on prod, prove it, and close with the written RCA.
A contractual 15-minute first response, Sundays and off-hours covered, front of the queue, and your access, alert channel, and runbook provisioned and drilled before any fire. A signed report every quiet month.
The full retainer for your 2–3 month peak window. Built for tax seasons, launches, and campaigns.
One fee per incident, and it covers everything through to the RCA. If the incident isn't resolved, you don't pay.
| Retainer client | Walk-up | |
|---|---|---|
| First response | 15 min contractual | Best effort, typically same day |
| Sundays + off-hours | Included | Available at twice the fee |
| Queue | Front, always | Behind retainer clients |
| Minute zero | Access and runbook provisioned, drilled | Starts with access requests |
| Incident fee | Same flat fee | Same flat fee |
Four years running production on AWS for a US tax-filing platform — first as an engineer, now leading the team — where every season the whole year's traffic arrives in one window. Teams across Chile, Colombia, Argentina, and the US, including two AWS consulting partners. AWS Certified Solutions Architect.
The ticketing system froze every peak day. The load balancer logs led to the heaviest request paths, to dashboard code looping over an internal API on every login, to an API behind it that could not scale. The symptom was in one system, the cause in another, the evidence in a third. Rewrote the code path, scaled the API, season saved.
SYMPTOM → LOGS → CODE → CAUSEA legacy application autoscaled onto a shared FSx cache, and latency never moved while AWS's own dashboards called the storage healthy. The saturation was in metadata operations, visible only after adding CloudWatch agent metrics and reading them against a load-test baseline. Moved the cache to NVMe instance volumes with stickiness. The system then held 170,000 requests per minute — 10,000 concurrent users in a 15-minute window — on architecture everyone had written off.
170,000 REQ/MIN · 10,000 CONCURRENT USERSThis is the entire service. Declare the incident at incidents@kernelpanic.pe and a senior engineer responds and gets hands on your system the same day, Monday–Saturday, 8am–10pm UTC-5. Retainer clients get a contractual first response and off-hours coverage.
A flat fee: $2,000 per incident, covering the firefight through to the written root-cause analysis. If the incident is not resolved, you do not pay.
No fix, no fee. What "fixed" means is agreed when you declare, before any work starts.
By AWS's own FAQ, AWS Support does not include custom code development, software debugging, system administration, or remote access to your systems. We log in and fix your system — infrastructure and application code together.
Yes: Peru, LatAm, and the US, in English and Spanish, on the US East Coast clock (UTC-5).
No staff augmentation, no permanent on-call, no open-ended "cloud optimization." After the RCA is delivered the engagement is over. Meetings beyond the two included calls are billed, visibly. Every engagement has a sharp question, evidence, boundaries, and an end.