A second opinion with the reasoning written down.
A technical decision review takes one costly, hard-to-reverse choice and gives it structure: the question, the evidence, the recommendation, and the alternative it rejects, from a reviewer who does not sell what comes next.
When to bring this in.
- The options will not resolve
- Build against buy, rewrite against modernize, platform against platform. Each side has a champion; the decision needs a referee with evidence.
- A vendor answer needs checking
- Proposals are written to be accepted. A review reads them to be tested.
- The stakes outgrow the room
- When a choice binds budgets and careers, an independent recommendation protects the people as well as the decision.
Decisions we review.
- Build versus buy
- Total ownership compared honestly: delivery, operations, exit, and the organization each option assumes.
- Platform and vendor selection
- Candidates read against constraints and lifecycle, not against their marketing.
- Rewrite versus incremental modernization
- What each path costs in releasability, risk, and time to first value.
- Cloud and infrastructure direction
- Where the estate should run, and what a move would actually change.
- Framework or architecture selection
- The choice weighed on lifecycle, team fit, and exit paths.
- AI feasibility and operating model
- Whether the case is real, what it depends on, and what operating it would demand.
What you receive.
- One question, answered
- The decision stated, examined, and resolved in writing.
- The evidence behind it
- What was examined and what it showed.
- A recommendation and its rejected alternative
- Stated with equal care, because the rejection is half the value.
- A decision record
- Reasoning your organization can revisit when circumstances change.
How it runs.
- Frame
- Agree the question and the constraints that bound it.
- Examine
- The systems, proposals, and organizations the choice touches.
- Compare
- Options against constraints, with the trade-offs in the open.
- Recommend
- One direction, argued; one alternative, rejected with reasons.
- Record
- The review document your organization keeps.
Bring the decision, not the conclusion.
Say what choice is on the table. The review starts from the question, not from a preferred answer. The proposal states the scope, the evidence required, the duration, and the fee before the engagement begins.