Building Velle: why we started

Every small architecture and engineering firm we talked to told us a version of the same story.

An RFP lands. Someone — usually the one proposal writer — drops everything to read it, dig through old proposals for relevant content, chase three sub-consultants for CVs and fees, and stitch it all together against a deadline that was already tight. The work that wins is relevance: the right past project, the right team, the right read of what the client actually wants. But most of the effort goes into finding and formatting, not into the thinking that wins.

The tools didn't fit

Enterprise proposal platforms exist, but they're priced for firms ten times the size — $15–25K a year, with onboarding to match. Generic document tools don't understand AEC at all: they don't know what a project sheet is, or why sub-consultant content has to stay attributed to the sub-consultant.

So small firms get neither. They make do with shared drives, email chains, and heroics.

What we believe

• Proposals are won by relevance, not volume. The software's job is to surface the right past content fast, then help write only the parts that genuinely need to be new.
• AI assists, humans approve. Every AI-generated element is flagged, sourced, and reviewable. Nothing ships without a person deciding it should.
• The platform respects existing tools. Word for editing, InDesign for layout, email for collaboration. Velle slots in — it doesn't try to replace your whole stack.

That's the wedge, and it's why we started. More on how it actually works in the posts that follow.