The release is the middle of the job, not the end.
Red Sentra takes over running applications and keeps them running: incidents, production defects, upgrades, deployment and the ongoing engineering a live system needs.
Responsibility
What we take on.
- Incident response, and the diagnosis afterwards
- Production defects, including the ones nobody can reproduce yet
- Dependency, runtime and platform upgrades
- Deployment, rollback and the release path itself
- Observability: the logs, metrics and traces that make the next failure legible
- The infrastructure the application runs on, where that is part of the engagement
Engineering
Maintenance and support are not the same word
Support is the response path: something is wrong now, somebody has to look, and the clock is running. Maintenance is the work that decides how often that happens — upgrades, dependency drift, the slow accumulation of things that were fine at the time.
A team that only does the first is always busy and never finished. Red Sentra treats both as one engineering responsibility, because the second is what makes the first rare.
This is not a helpdesk
It is not password resets, device support or a ticket queue staffed by people who cannot change the code. The engineers who answer are the ones who can read the service, change it and deploy the fix.
Production engineering is a mode of work here rather than an addition to a bill: the systems in this portfolio that shipped are systems the team still operates or still supports.
Keeping the evidence, not just the conclusion
A finding is a summary of some data, and a summary you cannot open is a claim. Cortex stores every probe exchange by content address and has claims reference the hash, so a wrong answer is diagnosable rather than mysterious.
The same instinct governs how a production system is instrumented. An incident report that cannot be reopened into the actual bytes is a story about an outage rather than a record of one.
Taking over something somebody else built
A handover starts with reading the system rather than rewriting it: what it depends on, what it assumes about the network, which parts have no owner, and which failures the team has learned to live with.
Modernising it, where that is the job, happens the same way — one boundary at a time, with the system live throughout. A rewrite proposed before anybody has read the thing is a guess with a budget attached.
Red Sentra owns the outcome — the application stays up and keeps changing — rather than supplying an engineer to sit inside another company’s rota.
Evidence
Systems this rests on.
Each entry names the contribution Red Sentra declared on that system. Nothing here is cited for work the content model does not carry.
RustSTT
A live service priced per audio hour, with partial transcripts during streaming, real-time-factor reporting and access control by domain or IP — operated rather than handed over.
Sentinel Solar
An always-on backend on a small computer at the installation, talking directly to the inverter and holding telemetry on site, where a maintenance visit is not a deploy.
NutriSense
A product on Google Play that Red Sentra owns and continues to operate, where an upgrade has to survive a device estate nobody controls.
Lexora
A commissioned product in production where Red Sentra keeps an ongoing support role rather than ending the engagement at handover.
Cortex
A content-addressed evidence store with deterministic, schema-versioned output, so a finding can be reopened into the exact response it was read from.
Limits
What this page does not claim.
- No uptime percentage, response-time target or service-level figure is published, because none has been agreed in public.
- No client of a maintenance engagement is named, and no incident belonging to a partner is described.