What should we prepare for the first project conversation?

You do not need a complete PRD. The following is enough to begin:

• the single most important problem and how it is handled now;
• who performs the work and its approximate frequency or volume;
• one normal, one common exception, and one difficult de-identified example;
• existing systems, data, devices, and third-party interfaces;
• an internal owner plus timing and budget constraints;
• legal, deployment, security, or platform conditions that cannot change.

When establishing communication, decisions, and change control, also compare Can we sign an NDA before discussing detailed software requirements? and Can we run a small pilot before committing to the full software project?; the linked guidance adds context that should be considered in the same decision.

During the conversation, Wavesteam first restates the goal and current process to expose misunderstandings. We then recommend whether mature SaaS, configuration and integration, custom development, or a bounded proof is the best next step and identify major risks and missing evidence.

When information is sufficient, the follow-up can include a direction, first-release boundary, assumptions behind an estimate, and the next action. When it is not sufficient, we state what discovery or technical proof is needed rather than promising a fixed total from one sentence.

Do not send material that cannot yet be de-identified. If deeper review becomes necessary, first agree confidentiality, named access, storage, and deletion. A useful one-line preparation is: “We currently use ___ to complete ___ about ___ times per month; the main problem is ___; in the first release, ___ users should be able to complete ___.”