Sentinel.
Live error sorter
See the problem once, not four hundred times.
One of fifteen tools the studio built for its own work. We build the same kind of app for clients, shaped around the job they need doing.
- Made for
- Keeping live sites healthy without drowning in noise
- Shape
- Native Mac app
- Data
- Stays on the machine
What Sentinel does.
Live sites produce a constant stream of error reports. Many are repeats; some show a customer unable to check out. Sentinel groups related reports into one problem and ranks it by how many people it affects.
One click creates a task for the team and tracks it through to completion. Our scheduled AI assistants investigate the grouped problem and suggest the likely cause for a person to confirm.
Find and rank problems before customers need to report them, replacing a long trawl through error records with a short daily review.
Why we built Sentinel.
Hand-sorting the error records was eating the start of every day. One underlying fault can appear as four hundred separate entries, and the customer quietly failing at checkout is somewhere in the middle of them.
Sentinel was built to make the sorting mechanical: recognise when hundreds of records are one problem, score it by how many people it touches, and put a short, ranked list in front of a person.
How Sentinel earns its place in a working day.
Reports arrive automatically
Error reports from every site we look after are collected continuously, without anyone watching a screen.
Grouped and scored
Duplicate records collapse into one underlying problem, scored by how many customers it touches. The widespread ones surface first.
The five-minute review
A heat-map calendar shows when problems spike; a ranked list shows what matters today. That's the whole morning check.
One click starts the fix
A task is created for the team, and our scheduled AI assistants suggest the likely cause for a person to check.
The morning check shrank
What used to mean scrolling raw error records is now a five-minute review of a short, ranked list.
Problems appear before complaints
Problems are found, grouped and ranked before a customer needs to report them.
Every fix leaves a trail
Resolutions are recorded with the change that fixed them, so the same problem twice is recognised, not re-investigated.
Made to fit the work it needs to do.
Sentinel uses fixed rules for grouping and ranking, giving the same answer each time. Our scheduled AI assistants then investigate the important problems and suggest the likely cause for a person to confirm.
We developed Sentinel using the client sites we already support. Its grouping handles variable details, its ranking reflects customer impact, and each problem can become a task someone owns.
Sentinel groups and ranks reports using fixed rules. AI then investigates the important problems and suggests a likely cause for a person to confirm.
Could your business use something like this?
Tell us the job in plain English. We'll say what would solve it, what it would cost, and whether a desktop app is the right shape for it.