FANG
(01)Work · SaaS / Design readiness platform

Dazzle.

Benchmarks a designer's portfolio, case studies and resume, then scores how ready they are to get hired, with a plan to close the gap.

Start a project
Dazzle: design readiness platform with benchmark score and readiness ring
Client
Own product · FANG
Year
2026
Timeline
9 weeks to launch · live 7 months
Role
Product strategy, Brand & UI, Full-stack build
Stack
Next.js · Postgres · Claude API · Stripe · Vercel
Dazzle Insights dashboard on a laptop: career signal, progress, rank and readiness signals

Designers don't know why they're not getting hired.

Portfolios get rejected silently. Dazzle's founders wanted a product that reads a designer's work the way a hiring manager does, and tells them, in one number and a short list, what to fix first. The hard part was making an AI judgement feel fair, specific and worth paying for.

  • ProblemNo feedback loop between applying and rejection
  • BetA readiness score designers trust more than a friend's opinion
  • ConstraintLaunch before the autumn hiring window

What we built.

Dazzle in four parts, the pieces that carried the product, in the order we shipped them.

  1. (01)

    Readiness score

    One number from 0–100 with a confidence ring, built from weighted benchmarks across portfolio, case studies, resume and LinkedIn.

  2. (02)

    Benchmark engine

    Structured rubric per asset, run through Claude with retrieval over our own dataset of hired-vs-rejected portfolios.

  3. (03)

    Fix-first plan

    Each weak benchmark expands into a prioritised checklist and before/after examples, re-scored on every upload.

  4. (04)

    Brand & site

    Warm, editorial identity that sits apart from the dark dashboards designers already ignore.

/ ScreensAs shipped
Two laptops: the learning path map and a lesson page
The path: a node map of what to learn next, and the lesson it opens
Visual Identity & Systems result: 83%, benchmarks and rank change
A benchmark result: 83%, where you sit against other designers, what moved

What it did.

Numbers shared as ranges; specifics on request.

  1. (01)

    Shipped in nine weeks

    Scoped to what proves the business: upload, score, plan, pay.

    Scope → launch
  2. (02)

    500+ users in seven months

    500+ designers scored in seven months, on web and iOS.

    Traction
  3. (03)

    Handed to a team

    Typed codebase, documented rubric, admin tools, no studio dependency after launch.

    Handover

Build something that ships.

Start a project
Squirrl: AI-native internal tool with voice orb, clients and documents
(02) Next projectSquirrlAI-native / Internal tool
All work