Brief Builder: Designing the Starting Point.
Team
Product Designer + PM + 2 Engineers
Timeline
Scope
Product design, research, prototyping, testing
Building an AI-powered tool that turned scattered documents into tender-ready briefs, unlocking CQuel's growth funnel.
CQuel connects real estate with sustainable solutions through bidding, but customers couldn't get started – brief creation was their biggest barrier. Without templates or guidance, potential projects never reached suppliers. We designed Brief Builder to transform this bottleneck into a guided journey, using AI to generate quality briefs from whatever documentation customers had.
01
The Problem
Customers need to create briefs before tendering projects. Most don't know how and get stuck.
Information Chaos
Data scattered across files. No idea what suppliers actually need.
Role Confusion
Who provides technical specs? Who handles strategy?
Status Uncertainty
Is this brief good enough to publish?
Manual Coordination
Endless emails chasing missing information.
02
The Research
Managers needed structure, not blank pages. They wanted guidance that helped them think, not just fill forms.
12
interviews
with managers
50+
existing
briefs
8
workflows shadowed
03
The Solution
What is Brief Builder?
An AI tool that helps customers create quality briefs without needing to know what to include. It guides them from scattered information to a confident "ready to publish" decision.
How it works
Customers upload whatever they have - PDFs, photos, notes. AI generates a structured brief with all required sections filled in.
Design principles:
Three core principles shaped how Brief Builder works
Direction over data
Show what to do next, not just what's missing.
Whose court clarity
Always visible who acts next and what they need to do.
Proactive status
he system tells them when the brief is ready.
Information Chaos → Structured Brief
Upload whatever you have - PDFs, photos, notes, scattered documents. AI analyses the content and generates an organised brief with all required sections filled in.
Role Confusion → Four Clear Ways to Fill Gaps
For each missing piece, choose how to complete it:
upload a document,
write it manually,
delegate to a team member,
or request help from CQuel.
No more guessing who should provide what.
Status Uncertainty → Decision Markers
Three quality indicators -
Pricing Accuracy,
Time in Tender,
Clarity - show if your brief is ready to publish.
Good/Excellent/Adequate ratings with clear "Ready to Proceed" signal replace guesswork with confidence.
Manual Coordination → Educational Guidance
Built-in explanations show why each piece matters:
"Without structural docs, contractors add 15% safety margin" or "Visual confirmation reduces electrical estimates by 10%."
Customers learn tender dynamics while building their brief.
04
Testing
What we did:
Built a fully functional prototype using Cursor - not just clickable, but genuinely working like a real site.
Tested with 5 customers (asset managers, consultants, engineers).
CEO used it for sales demos while we were still designing.
What changed and why:
Percentages → Simple Ratings
Users didn't want scores - they wanted to know "is this good enough?" Changed to Good/Excellent/Adequate with clear "Ready to Proceed" signal.
Our Jargon → Their Words
Terms like "best bids" confused people. Switched to plain language customers actually use in their work.
Everything in One View → Separate Tabs
Editing and reviewing at the same time caused confusion about what state they were in. Split into Review and Optimise.
Text Editing → Recommendation Actions
Users wanted to trust the AI baseline, not constantly question it. Removed inline editing.
05
Impact
Validated foundation for Deal Closing Management - next product phase.
Core insight:
Customers need clarity on "whose turn is it?" This extends beyond briefs to entire project lifecycle.
06
Key Learnings
Validated foundation for Deal Closing Management - next product phase.
Functional prototypes enable sales
Cursor prototype became a sales tool before dev shipped.
Design for teams, not individuals
B2B workflows need delegation built in.
Use customer language
Their words = instant understanding.
Show direction, not just data
Interpretation matters more than information.
Previous












