Operating Model
Five stages. One direction of travel.
Each stage has clear entry criteria, a required artifact, and a decision gate. The model is linear at the seams and iterative inside each stage.
01
Intake
- Answers
- Q1 · Which problem is worth solving?
- Entry
- A named sponsor and a measurable business problem.
- Artifact
- Signal brief with problem frame and first hypothesis.
- Gate
- Triaged: pursue, park, or route.
02
Prototype
- Answers
- Q2 · What should we prototype?
- Entry
- Pursued signal with a testable hypothesis and target users.
- Artifact
- Working prototype in front of real users.
- Gate
- Shareable prototype ready for validation.
03
Validation
- Answers
- Q3 · How do we validate it in a week? · Q4 · How do we get executive buy-in?
- Entry
- Prototype and a named riskiest assumption.
- Artifact
- Evidence pack + decision recommendation.
- Gate
- Iterate, pivot, pause, or invest.
04
Readiness
- Answers
- Q5 · How do we know whether to invest?
- Entry
- Validated bet with an invest recommendation.
- Artifact
- Readiness report: security, scale, ops, cost, governance.
- Gate
- Cleared for handoff — or gaps named.
05
Handoff
- Answers
- Q5 · How do we know whether to invest? (final call & transfer)
- Entry
- Cleared readiness and owner signed on.
- Artifact
- Handoff package + 30/60/90 support plan.
- Gate
- Owned by the receiving team.
Guiding principles
How we make decisions inside the model
Prototype before you debate
Working software creates better conversations than presentations.
Evidence over opinion
Decisions are supported by customer feedback and measurable learning.
AI first
Assume AI can assist every step unless there's a reason it shouldn't.
Build to learn
Every prototype exists to answer questions and reduce uncertainty.
Enterprise-ready thinking
Even early concepts consider security, governance, and operational impact.
Continuous improvement
Every engagement improves the playbooks, templates, and tooling.
Where this model ends — and Forward Deployed Engineers begin
Clarity before commitment. Commitment before code.
The five stages exist to turn a fuzzy signal into a validated bet with evidence. That's the seam Forward Deployed Engineers can't afford to work in — sending senior engineers on unvalidated fishing trips is expensive discovery in disguise.
Intake → Validation
Accelerators
Answer whether to build and what exactly to build. Prototype-led learning, no commitment yet.
Readiness
Joint
FDEs can be pulled in as Supporting Partners to sanity-check feasibility, scale, and ops before the invest call.
Handoff → Production
Forward Deployed Engineers
Take the evidence pack + readiness report as their brief. Ship the committed bet in production, embedded with the customer.
See the model in action
The Portfolio shows every live bet by stage — the operating model at work today.