Case studies

Every case study here follows the same structure, including the part most agencies leave out: what didn’t work first time.

Straight answer: there are no case studies here yet, and we won’t invent any. Every deployment we publish will name the workflow, the systems, the measured before-and-after — and what didn’t work first time. Until then, the audit is how you get evidence about your own business rather than someone else’s.

Case studies populate as engagements complete

No engagements have been documented yet. This page is intentionally empty.

When case studies are published, each will follow the same structure: the workflow, the systems involved, what was built, the measured results, and what didn’t work first time.

The structure every case study follows

This is the template, published so you can see what we will and won’t claim. It stays here until there’s a real engagement to put through it.

[Outcome-led title — lead with the number, not the client name]

Industry: [ ] · Systems: [ ] · Build time: [ ] · Live since: [ ]

Why existing tools fell short

Open with the gap, not the client bio. What did they already own, and why didn’t it solve this? This is the strongest opening pattern in the category — it establishes that the problem is real before you claim to have solved it.

The workflow

What the work actually looked like before. Who did it, how often, how long it took, where it broke.

What we built

The agent. Which category, which triggers, which systems it reads and writes, where it escalates.

Key design decisions

[Decision 1] — what we chose and why. Include the alternative you rejected.

[Decision 2] — as above.

Shadow mode: what we learned

What the agent got wrong during validation, and how thresholds changed as a result. Include this. It is the most credible section in any case study and almost nobody writes it.

Results

Metric	Before	After	Measured over
[ ]	[ ]	[ ]	[ ]
[ ]	[ ]	[ ]	[ ]

Only numbers we can defend, with the measurement window and the method cited.

What we’d do differently

Genuine. One or two things.

What’s next

The workflow after this one, and whether it’s been scoped.

Frequently asked questions

Why are there no case studies yet?

Because we won’t invent them, and buyers in our verticals talk to each other. Every study published here will be real: named workflow, named systems, measured results, and the part most agencies omit — what didn’t work first time.

How can I evaluate you without case studies?

Three ways that don’t require trusting us: the published stack on /how-we-build/, the live demos, and the audit itself — a fixed-price deliverable you keep whether or not you build with us.

Can we be a case study?

Early clients who agree to a named, verifiable study get preferential terms. Ask during the audit.

Nothing here yet that matches your industry?

Case studies accumulate. If yours isn’t represented, the audit is designed exactly for that situation — you get evidence about your own business rather than someone else’s.

Book the 5-day audit →See all industries →

Portfolio of Shipped AI Agents