You sell hours.Admin spends them for free.
In a studio every hour is either billable or gone. The work worth automating is not the craft — it is the scoping, status, timesheet and invoice layer that surrounds it and bills to nobody.
Open source projects
Public on github.com/JIGGAI
AI plugins
Published on npm, free to install
AI plugin installs
66,018 to date, per npm
AI monthly plugin installs
2,811 in August 2026, per npm
What we hear in the first hour.
Not a generic list. These are the ones that come up in businesses shaped like yours.
Timesheets get filled in on Friday from memory.
Which means they are wrong, and under-recorded rather than over. Recovering that is usually the largest single number in the report.
Scoping a project takes a week of senior time and half of them do not land.
The unsuccessful half is unrecovered cost. Drafting from previous scopes turns a week into a day without touching the judgement.
Clients ask for status and someone stops working to write it.
Status exists in your project tool already. Assembling it into something a client will read is the manual step, and it repeats weekly.
We only find out a project is underwater at invoicing.
Burn against scope is knowable weekly. Nobody watches it because watching is unbillable and therefore always deferred.
Everyone here has already tried AI on their own and nothing stuck.
Individual tools versus a studio process. What did not stick was the standardisation, not the model.
Your product is judgement sold by the hour. So every unbilled hour is the whole margin.
That makes this sector unusual: the return is not efficiency in the abstract, it is hours moved from the unbillable column to the billable one, and both columns already exist in your own system.
The software already in the building.
We do not arrive with a platform to sell you. We arrive expecting these, and the work is usually in the gaps between them.
Project and time tracking
Harvest, Float, Productive, Deltek. Correct only to the extent people fill it in honestly and promptly.
The scoping spreadsheet
One per principal, each with their own rate assumptions. Where pricing knowledge actually lives.
Project management
Asana, Notion, Jira, Monday. Where status exists in a form no client would read.
Client communication
Email and Slack Connect. Where scope creep is agreed to informally and never makes it back to the scope.
File and asset storage
Drive, Dropbox, or a server. Where finding last year’s equivalent job takes longer than redoing part of it.
Accounting and invoicing
Xero, QuickBooks. Downstream of the timesheets, and therefore as wrong as they are.
Making the timesheet a review rather than an act of memory.
Roughly what one entry in your report looks like — a real shape, with the numbers changed.
- What we saw
- Timesheets completed weekly from recollection, well after the work. Studio-wide, recorded hours ran consistently below scheduled hours, and nobody believed the difference was idle time.
- What it costs today
- The gap reads as unrecorded billable work — worth more, on any conservative reading of it, than any efficiency saving available elsewhere in the building.
- What we would build
- A daily draft assembled from what the person already touched — calendar, project tool, file activity — presented as a pre-filled timesheet to correct rather than a blank one to fill. It stays a draft until a human confirms it, and the corrections train what gets suggested.
- How you would know it worked
- Recorded hours against scheduled hours, and how long after the work the timesheet is confirmed. If people accept the draft without reading it, that is a failure mode and we would rather find it than not.
Submitting it. A timesheet is a claim about what someone did, and a system that filed it unread would be putting words in their mouth to a client — and eventually to a dispute. The confirmation stays a deliberate human act.
What coordination work costs you.
Three numbers you already half-know. Move them until they look like your business.
re-keying, chasing status, producing the same document again
on that work specifically, not their whole job
salary, tax, benefits, desk
A 75% capture rate. The other twenty-five per cent is judgement, exceptions, and not wanting to look greedy.
This is arithmetic, not a finding. It rests on three numbers you guessed. The assessment replaces all three with numbers we observed — and tells you which of those hours are actually worth automating.
See what a real finding looks like →What the six days look like.
The most common question we get is not about AI. It is what these people will actually do in my building.
- Day 1Walk the floorWhoever is on shift
We start where the work happens, not in a meeting room. Nobody prepares anything, and the first day is mostly watching.
- Day 2Sit with the people doing itOps, admin, finance
Conversations with the roles that touch the work most, and a real task followed end to end — including the parts that happen in a group chat.
- Day 3Systems, then a read-backWhoever holds the logins
What you run, what talks to what, and where a person is currently the integration. We tell you what we saw before we leave, while it is still cheap to correct.
- Days 4–6Research and discoveryOur desks, not yours
Away from your building. We cost the work we watched, model the alternatives, and test the shortlist against your own numbers rather than a framework.
- +1 weekOne recommendationPresented in person
The single change worth making first, specified precisely enough to build — with the ranked analysis behind it and the list of what we would not automate.
The questions this raises.
Will this end up monitoring our people?
That is the risk, and it is why the recommendation is a draft the person corrects rather than a record of their activity. If a finding could only work as surveillance, it goes on the do-not-automate list — a studio that feels watched loses more than the hours it recovers.
Our work is creative. Is any of this relevant?
The craft is not what we look at. The scoping, status, recording and invoicing around it is identical to every other professional service, and it is where the unbilled hours are.
We are twelve people. Are we too small?
Twelve is fine — five is where we start. Small studios often have the clearest findings, because one person can explain the whole operation in an afternoon.
Everyone here already uses AI tools individually.
Which is useful information and usually part of the finding. Individual adoption without a studio-level process is why nothing compounds: the same problem gets solved five times, differently, and none of it survives the person leaving.
Start with half an hour.
We'll tell you whether an assessment would pay for itself in an operation like yours. Sometimes the answer is no.