Track 10 · Discovery & handoverLesson 07 of 12
Running a bootcamp on live data
The workshop that proves value and sells the next phase.
- Lesson library · reference
- 35 min read
- 60 min hands-on
- Halden Freight
- Pro
What you'll be able to do
01- Explain the three purposes of a bootcamp (enable the customer's team, prove value on their data, find blockers fast) and why it is not a training course, a hackathon or a demo.
- Design the pre-work for a bootcamp: access, sandbox, scoped data, pre-qualified use cases, signed success criteria, people and security sign-off, each with a named owner and a due date.
- Implement a readiness checker that computes days at risk, returns a go / reshape / no-go verdict, and generates an agenda and readout template from the selected use cases.
- Run a 1–5 day agenda of build-measure-demo loops on access-scoped live data without writing to production.
- Facilitate a day-5 readout that ends in a recorded decision, and follow it through so the momentum survives the following Monday.
Before you start
- Scoping a shippable slice: the bootcamp builds the slice. Without a slice it becomes a hackathon.
- Feasibility and data access, fast: the access model the bootcamp runs on.
- Baselines and ROI you can defend: the numbers the readout compares against.
- From requirements to eval criteria and Golden datasets and the metrics that matter: success criteria and eval seeds.
- Finding and masking personal data: how the data is made safe to touch.
The short version
A customer bootcamp is a time-boxed, hands-on build, usually one to five days. The customer's own people work on their own access-scoped live data and finish with a working artefact, results measured against criteria agreed in advance, and a recorded decision. It is won or lost in the pre-work: if access, sandbox and success criteria are not ready before day 1, you are paying ten people to wait. Each day runs build-measure loops and ends in a demo that the customer team drives. The last day ends in a readout and a decision with a name and a date on it. Then you follow through within 48 hours, or the energy disappears.
- No access, no bootcamp. Day-1 items are gates, not intentions. Slip or reshape rather than start blind.
- The customer drives the keyboard. You coach and unblock. Their team presents on day 5.
- Live data, scoped. Read-only, masked, sandboxed, drafts only, no production writes.
- Every day ends with a demo on real data and a captured artefact. Eval seeds, prompts, tool specs, backlog and risks are deliverables.
- The readout ends in a decision. Proceed, iterate or stop. Applause is not an outcome.
Pro lesson
The rest of this lesson is in Pro.
The mental model, worked example, field note, failure table, exercise and self-check, with the voice tutor.
- Every stage and specialisation, in full
- Every lab and every role project
- Halden weeks 2–10, on your record
- Every drill, live in the playground
- All solutions, plus the server judge
- Lab Assistant and every Teach with AI