AI workflow consulting
We find the work in your business that AI can do well. Then we build it into the tools your team uses now, with checks that keep a person in control.
Talk about your project§ 01What you get
The result, in plain terms.
- A short list of tasks where AI saves real time, with an estimate for each.
- Automations that run in your current tools, not a separate demo.
- Clear rules for data, privacy and human review.
- Measurements that show what each workflow saves.
- A team that knows how to run, change and stop each workflow.
§ 02The route
Stage 1
Survey
We talk to the people who do the work. We map each task, its inputs, and the cost of an error.
Stage 2
Choose
We score each task for value and risk. You choose the first two or three to build.
Stage 3
Build
We build each workflow with your data and tools. A person approves the output where an error costs money or trust.
Stage 4
Measure
We record time saved and error rates for each workflow. We remove a workflow that does not pay for itself.
§ 03Ways to start
| Engagement | Scope | Typical length |
|---|---|---|
| Opportunity map Interviews, a task map and a ranked list of AI opportunities with costs. | Interviews, a task map and a ranked list of AI opportunities with costs. | 2 weeks |
| First workflows Two or three workflows in production, with review steps and measurements. | Two or three workflows in production, with review steps and measurements. | 4 to 8 weeks |
| AI policy and governance A usage policy, a risk register and a review process. This work can align with ISO/IEC 42001. | A usage policy, a risk register and a review process. This work can align with ISO/IEC 42001. | 2 to 4 weeks |
§ 04Questions
Which AI tools do you use?+
We choose for each task. For some tasks, we use hosted models such as Claude. For other tasks, a self-hosted model costs less and keeps your data on your own servers. We also use the automation tools that you already pay for. We do not resell software.
Is our data safe?+
We use business plans that do not train on your data. For sensitive data, we can run a self-hosted model, so the data does not leave your servers. We write down where each piece of data goes.
Do we need a technical team?+
No. We build workflows that your staff can operate. We also train one or two people to own them.
Next step