Contact

Start with the problem, not the pitch.

Describe the work, the current failure mode and the evidence a successful system would need to produce.