Who it’s for: Owners and managers of businesses with 20–200 staff and no in-house tech team
Your fractional AI team
A monthly retainer that works through your AI backlog one process at a time: pick, build, test, train and measure. It runs after a discovery.
The problem
Large companies have internal AI teams. They collect ideas, rank them, build the best ones, train staff and measure what happens. Growing businesses rarely have that, so AI stays a pile of tool trials: a few subscriptions, a few enthusiasts, and nothing that anyone owns or keeps working.
What we build
- A monthly cycle that works through your backlog, taking on the next process each month
- AI built into that process, tested against your own examples before anyone relies on it
- Guardrails, monitoring and staff training, so the work keeps running after we move on to the next one
How it works
Pick the next process
At the start of each month we choose the next item from the backlog with your owner, based on value, risk and effort. Our business analyst keeps the backlog current as priorities change.
Build and test
Our AI engineer builds the system and tests it against examples from your own work, with a pass mark agreed with the person who normally does the task. A person reviews the output until you are comfortable.
Train and hand over
We train the people who will use it, write a short runbook and agree who to call when something looks wrong. Your champion in that team helps us get adoption right.
Measure and report
Each month you get a short report: what shipped, what is in progress, how much it was used, any issues, what it costs to run and what we suggest picking next.
What it looks like
Illustrative example of a backlog for a fictional wholesale business. Supplier invoice checks (high value, small effort) are live, customer enquiry triage (high value, medium effort) is in build, a weekly sales summary (medium value, small effort) is next, and a staff policy assistant (medium value, medium effort) is later.
What you receive
- Working AI in one process at a time, connected to the systems you already use
- A monthly report covering shipped work, work in progress, usage, issues, running costs and the next pick
- Guardrails and monitoring for everything we have built
- Staff training and a runbook for each process
- An up-to-date backlog, ranked by value, risk and effort
How we check accuracy
- Every build is tested against examples from your own work before it goes live.
- A pass mark is agreed with the person who normally does the task, and a person reviews the output until you decide otherwise.
- The monthly report shows usage, issues and running costs, so you can see what is working and what is not.
What changes
- AI working in one real process at a time, instead of a pile of trials
- A backlog that someone keeps ranked and up to date
- A monthly report that shows what shipped, what it costs to run and what comes next
What stays with you
Decisions about what to build, who may use it and when to rely on it. You name one owner who sets priorities, and one champion in each team we work with.
In practice
Think of this as the AI team a larger company would hire, shared across a few clients and run as a monthly service. It is not a licence for a product. It is people who work through your list, one process at a time, and leave something that works and that your team knows how to use.
Our business analyst runs the backlog and the adoption work. Our AI engineer builds and runs the systems. You name one owner who decides priorities, and one champion in each team we work with. That is all the structure we need from you.
Each month follows the same cycle, so you always know what is happening. We choose the next process, build and test it, train the team and report back. If a process turns out to be better left alone, we say so and move to the next one.
Questions
How is the retainer priced?
Quoted after discovery, based on how much we take on each month.
Can we pause or stop?
The terms for pausing or stopping are agreed in your client agreement.
Do we need a tech team?
No. We need one person who can decide priorities, and one champion in each team we work with.
Who does what?
Our business analyst runs the backlog, the training and adoption. Our AI engineer builds and runs the systems. You name an owner and a champion for each team.
Do we have to start with discovery?
We start every engagement with discovery. It gives us the backlog and tells us both whether the retainer is a good fit. The discovery fee is credited toward the retainer.
Where does our data go?
For client engagements we build on AWS in Australia (Sydney region where available); any exception only with your written agreement.
Related
AI discovery
A fixed-fee AI discovery, credited toward your retainer. We map how your work runs and rank where AI could help first.
Find where AI helps first
AI document processing
Reads supplier invoices, purchase orders, forms and contracts, checks them against your rules and flags exceptions for a person to decide.
Read, check and flag documents