Manage client sites like an agency, even when it is just you.
Unolia watches every client site the way a bigger team would. It finds what silently broke, prepares the fix and keeps the story per client, while your hours go to the work you bill.
YouOne personEvery client covered
fermeduport.frwatched
galerie-ondine.frfixed today
cabinet-morel.frwatched
Monitors green · 0 open issues
Unolia does the watching
Works with the providers your clients already useAmazon Route 53BentoBunny.netCloudflareDigitalOceanGandiGitHubGitLabIonosLaravel CloudLaravel ForgeMailgunNamecheapOh DearOVHcloudPloiPorkbunVultr
The real problem
The watching is a full-time job. It is not yours.
An agency puts someone on maintenance. Alone, the checking waits for the hours after client work, and those hours rarely come. These are problems Unolia caught without being asked.
no DMARC→record added
rechecked green
The problem you did not know to look for.
The first sync on a domain you had managed for years flagged something no client ever asked about: no DMARC record, on a domain that sends invoices every day. The issue explained what that means in plain words, the fix came attached, and applying it took one click. Knowing about DMARC was never the job. Now it does not have to be.
A client's site moved to a new host on a Friday. One DNS record kept pointing at the old server, so part of their visitors still landed on the stale site. Unolia flagged the mismatch, proposed the repoint, and the recheck came back green before dinner. No forum threads, no command lines, no weekend.
Somewhere, an advisory was published for a package one client site depends on. That day, you were on another client's launch. Unolia checked the repository against it, matched the vulnerable version and opened an issue on the project. You shipped the update on your schedule, not after an exploit.
The same week you already run, with the watching and the reporting done for you.
Monday 08:40
The question answered before the coffee is cold
A client reads about a hack in the news and emails to ask if their site is fine. You open the project: monitors green, no open issues, the last deploy two weeks ago. The reply takes two minutes and reads like it came from a shop with a support desk.
answered at 08:42
Wednesday 07:30
The maintenance run you start with one click
The pre-made run was configured once. This morning you press the button with your coffee: updates across all 5 client servers, a reboot for the one machine that asks for it, every step logged under the right project. You read the result instead of connecting to five servers one by one.
5 servers, 1 reboot, all logged
Thursday 18:20
The colleague you can finally ask
One client runs on a stack you touch once a year. When it misbehaves, you ask your AI assistant. It reads the project Unolia keeps, pulls the site's brief and its environment map, and returns the likely cause with a fix to review. Working alone stops meaning figuring out everything alone.
likely cause, fix to review
Friday 17:00
The retainer email that writes itself
The monthly recap used to be the chore that slipped. Now the project already holds the month: issues caught, fixes applied, run logs, what was spent. You copy the story out instead of reconstructing it, and the client sees what a quiet month actually contains.
the month, already on record
Watched all week, even when it is just you
Your AI assistant
An assistant that already knows every client site.
An agency has a colleague to ask. You have an AI assistant with Unolia behind it. Connect the tool you already use, Claude Code, Cursor or VS Code, and your assistant reads your client projects through the Model Context Protocol (MCP). It answers from what Unolia knows, not from guesses.
It can brief you on any client site, explain an incident, read the run logs, list what renews next month and apply a supported fix after showing you exactly what will change. You stay the one who decides. It becomes the one who looks.
Things you can just ask
“What is the state of fermeduport.fr?”“What renews across my clients next month?”“What did this morning’s run change?”“Add the missing DMARC record, show me first.”
The monitor says galerie-ondine.fr is down. What changed?
The 17:52 deploy changed the cache config. The monitor went red two minutes later. Here is the deployment and the log line to check.
And what renews next month?
Three renewals: the fermeduport.fr domain at Gandi on the 12th, an SSL certificate on the 18th and one hosting invoice at the end of the month.
read from the project, not guessed
The value
Where the retainer earns its price.
The retainer stops being invisible.
Maintenance is real work that nobody sees when it goes well. The project keeps the month on record: issues caught, fixes applied, runs completed, money spent. When the renewal conversation comes, you show the story instead of asserting it. Clients renew what they can see.
Last month, on record
4 issues caught
3 guided fixes applied
2 maintenance runs, fully logged
€84.50 provider spend, allocated
pasted into the renewal email
Know what each client really costs you.
Provider bills, manual entries and uploaded invoices land on the client they belong to. Renewals show up before the card is charged. When you price the next retainer, the number comes from what the client actually costs you, not from a feeling.
fermeduport.fr€41.00
galerie-ondine.fr€23.60
cabinet-morel.fr€19.90
1 renewal aheadinvoice uploaded
Proof
The product behind the stories.
Every scene above is a capability included in the Freelancer plan, not a promise on a slide.