Forward-deployed delivery

The two-week Discovery Sprint, unbundled.

Inputs, outputs, kill criteria, and how Brandlabs maps a Discovery Sprint into a 12–20 week zero-to-PMF engagement.

HV
Harsh Vitra

·7 min read

TL;DR

A Brandlabs Discovery Sprint is two weeks to pressure-test thesis, data, compliance posture, and eval targets — then make an honest go/no-go call. It is not a free mini-build. The artifact is a plan you could hand to another team and still respect.

What is a Brandlabs Discovery Sprint for?

To decide whether a zero-to-one AI product can be shipped in a fixed window with the data, constraints, and team you actually have. Success is clarity — including a justified kill — not a prototype that papers over missing BAAs or empty eval criteria.

We meet your team, your data reality, and your constraints. We leave with:

  • Architecture options with tradeoffs
  • Compliance sequencing (HIPAA / DPDP as applicable)
  • Eval dimension targets
  • Risk register
  • Recommended engagement shape (or a no)

If the only acceptable output is “yes, build,” you do not want a discovery. You want affirmation.

What inputs do you need before week one?

A product thesis, access to decision-makers, sample data or a clear path to it, known vendors, and any hard deadlines or regulatory constraints. Without those, the sprint becomes interviews that cannot falsify the plan.

Minimum useful kit:

  • Who the user is and what job the AI does
  • Where data lives today
  • What “unsafe” means in your domain
  • Budget and timeline honesty
  • Named owner on the client side

We will not invent PHI access you do not have. That fiction destroys the 12–20 week plan.

What are the kill criteria?

Kill or reshape if data is inaccessible, BAAs/vendor posture cannot support the thesis, success metrics are undefined, or the timeline requires magic. A clean no in week two beats a muddy yes that fails in week fourteen.

Examples of kill/reshape signals:

  • “We’ll get the EHR feed eventually” with no owner
  • Consumer LLM path required for PHI
  • No one will own eval thresholds
  • Scope is “an agent that does everything”

Brandlabs would rather lose the build than win a fantasy SOW.

Frequently asked questions

Is the sprint free?

No. It is a paid, time-boxed engagement with a real deliverable. See the contact page to apply.

Do we always continue into a pod after discovery?

Only when the plan is honest and mutual. Discovery is designed so both sides can walk.