Illustrative example

Discovery notes into a reviewable proposal

A proposed drafting assistant that connects discovery notes with an approved scope library while leaving commercial decisions with the team.

DiscoverDraftReview

A proposed workflow, not a completed client project. No client, deployment or measured result is claimed.

The example brief

In this illustrative agency, useful discovery notes exist but proposal drafting starts from a blank document. Details about the client’s tools, desired outcome and constraints are copied several times. A sensible first scope is a structured draft for one repeatable service, with a separate commercial review before anything reaches the client.

The workflow we would scope

We would organise the notes into the problem, current process, proposed deliverables, dependencies and open questions. The draft would draw on an approved service library and retain the source of important requirements. Missing facts would appear as questions for the author. Prices and terms would come from the approved commercial source or remain explicitly unfilled.

Where a person stays involved

The account owner would decide scope, effort, fees and commitments. An assistant should not invent a delivery date to make the document look finished. Conflicting notes would be highlighted rather than silently reconciled. The final version would require a named reviewer, and external sending would remain a distinct authorised action after that review.

What a pilot would need to prove

Test sparse notes, contradictory requirements, a request outside the service library and notes containing irrelevant instructions. Check whether the draft preserves exclusions and access dependencies. Compare the approved document with the underlying brief. Confirm that later changes invalidate the previous approval and that a draft cannot be mistaken for an accepted contract.

How success would be measured

Measure drafting effort, reviewer edits, omitted requirements and time waiting for missing information. A faster draft is not automatically a better proposal, so review accuracy alongside speed. This example claims no win-rate improvement. A useful pilot would retain reviewer feedback and compare proposals of similar complexity before deciding what to automate next.

Further reading

These links provide context for the proposed workflow; they are not evidence of a completed client engagement.

Your next project

Start with the problem
you want to solve.

Tell us what happens today, where the work gets stuck and what a useful first result would look like.

Discuss your project

No technical brief needed.