Restraint over coverage
Instrumenting everything would have been easier. The useful version explains the failure without becoming another thing the estate has to trust blindly.
Case study · AI & Cloud
Ops-Intelligence
See the stack, not just the symptom.
Ops-Intelligence watches enterprise apps.
Read without intruding
Ops-Intelligence. Product interface.
No hooks
logs already on the machine
Rules first
the model only sees what is new
Operator
nothing changes state on its own
The challenge
A desktop is a poor witness
The visible symptom points at the wrong machine, and another log file does not correct that. The work was to correlate what the estate already records, across the services that depend on each other, without hooks into the running system.
The usual agent - invasive hooks, binary changes, a stack of third-party dependencies - is what gets this kind of tool rejected. A passive sense layer reads what the endpoint already exposes and builds a graph. What gets reported is the dependency that explains the failure, not the process that happened to be in front.
Built by Maatara Consulting
Engineering case study. Implementation details and outcomes depend on the deployment environment.
The engineering approach
What this work asks of a team
From the wrong machine to the real cause
Cost as a design choice
Calling a model on every event looks intelligent and behaves expensively. The rules carry the common path. The model is a specialist.
A change that can wait
Proposing a fix and applying it are different jobs. Keeping them apart is what lets this sit in production rather than in a demo.
The stack this product is built on.
- Python
- Go Lang
- React
More work
More AI and Cloud work
SecBoard
The command centre for intelligent vulnerability management.
- C#
- MySQL
- Go Lang
Digital Signage
Run displays anywhere, manage everything from one portal.
- C#
- SQL Server
- Flutter
Need something like this built?
Tell us what you are trying to ship and an engineer will reply within one business day.