Build the proposal around your expertise, not guesswork.

The question behind the proposal

She knew her work and was already putting proposals together. What she wanted was a place to think: Does this approach fit the client's problem? Is there a useful comparison with how other consultants structure similar work?

Those are different questions. One needs a thinking partner. The other needs evidence. Treating both as a request for a quick AI answer risks adding polished guesswork to a proposal that already has a sound foundation.

Start with the client, not the search results

We designed the process to begin with what she already knew about the client and what she believed belonged in the proposal. The assistant offers a working hypothesis first, with a reason behind it.

Research has a specific job: check outside practice when the proposal depends on it, resolve an uncertainty that evidence could settle, or answer an explicit market-comparison question. A broad question about whether an idea is good stays a thinking conversation until there is a reason to look outside.

What we built

• ◆A proposal thinking partner that starts with the consultant's own understanding of the client's problem.
• ◆Targeted research when an outside comparison would help, with client names kept out of public searches.
• ◆A requirement that claims about other consultants come from sources retrieved in that conversation, not a plausible-sounding answer from memory.
• ◆One focused question when a missing fact would change the proposal, rather than a long intake before useful work can begin.

What stayed human

The consultant owns the recommendation, the scope, the price, and the relationship. The assistant does not invent a fee, a deadline, a name, or a commitment to fill a gap.

The email that accompanies the proposal is a separate step, in her voice. Reviewing and sending it remain hers. Help thinking through the offer is not permission to speak for the person making it.

What this makes possible

A useful proposal review does not have to find something wrong. If reliable sources support the approach, the process is designed to say there is no reason to change it. If the evidence is weak, it says that rather than manufacturing a comparison.

The value is a more grounded way to think through an offer: a recommendation to consider, evidence where it belongs, and a clear question where judgment is still needed. The consultant can test her thinking without letting a search result become the strategy.