Skip to main content

PortfolioCase study

VL-01Live · mock data

Nuptis → Velora

Where brands and makers find their fit.

Nine days.Two products.One survived.

A B2B apparel sourcing marketplace where brands and manufacturers match, then turn matches into bids.

  • 2

    products

    Structural

    Nuptis, then Velora — nine days, solo

  • Day 7

    first product killed

    Structural

  • 1 day

    to build Velora

    Structuralas of 11 Aug 2026

Evidence keyMeasuredStructural

The killed first product, Nuptis: a vendor roster dashboard for a wedding-planning agency, with mock vendors, risk tiers and compliance status.
Velora trust profile for a mock manufacturer, Loomcraft (organic knits and jersey, Tiruppur): a 94 out of 100 score labelled High Trust.
Trust score · authored prototype data
Velora's first screen on a phone: “Where brands and makers find their fit. Trust-first apparel sourcing — swipe, match, bid.” with an “I'm a Brand” card.

The problem

Supplier trust is found by asking around — and it doesn’t travel.

An indie apparel brand needs a garment maker; a factory in Tiruppur or Ludhiana needs brands.

How a brand finds a factory today
  1. Cold referrals
  2. Trade fairs
  3. Directories
  4. Unverified trust
  5. Trust that doesn’t travel

Research (team)

The delay is waiting, not work.

Procurement interviews were pooled across the Case Study 3 team.

Where onboarding time goes

15–30 business days

Active work · under 10%Waiting between hand-offs

Team secondary research, marked “verify before external use” — shown as background, never as Velora’s data.

  • “I find out where a vendor is by asking around.”

    Procurement interview, team PRD
  • “We scrutinise new vendors. Changes to old ones, we just… trust.”

    Procurement interview, team PRD

The pivot

Kill the first bet on day seven.

  1. Days 1–7

    Nuptis

    Vendor ops for wedding-planning agencies — live, with every success metric defined and none measured.

    Day 7 · killed

  2. The evidence

    “Weddings were blue — but a shallow pool.” Few events, low willingness to pay.

  3. Keep the insight

    Keep the trust-and-onboarding insight; aim it at apparel sourcing.

  4. By day 9

    Velora

    Brands and manufacturers match; matches turn into bids. Built in a day.

The bet

Portable trust.

One trust profile a manufacturer carries into every match, RFP and bid. The scores are authored prototype data — verification was out of scope.

BrandManufacturer
  1. Match
  2. RFP
  3. Bid
Velora trust profile for a mock manufacturer, Loomcraft (organic knits and jersey, Tiruppur): a 94 out of 100 score labelled High Trust.

Trust score · authored prototype data

Trust profile
Velora “Your RFPs” screen listing two live mock requests, Organic Cotton Tees and Linen Shirt Run, with units, budget and ship-by dates.
A brand’s RFPs
Velora “Bids received” screen: mock manufacturer bids on an Organic Cotton Tees RFP, sorted by best match, each with price per unit, lead time and trust score.
Bids, sorted by best match

Key learnings

Nine days. Two products. One survived.

  • Kill without flinching

    A blue ocean that is a shallow pool is still the wrong ocean.

  • Attack the hand-offs

    Onboarding delay is queue-time between teams, not any one team working slowly.

  • Label honestly

    Mock data isn’t a proven path, and an authored score isn’t a verified one.

Evidence

A working prototype — not traction.

  • 10/10

    unit tests passing

    Measuredas of 15 Sep 2026

  • 0

    horizontal overflow

    Measuredas of 11 Aug 2026

    every route, at 375 and 768 px

  • 1 day

    build time

    Structuralas of 11 Aug 2026

    all 40+ build commits on 11 Aug 2026

Not measured

  • Real users
  • A real pilot
  • Usage data
  • A verified trust score

Mock data

Source material

Everything here is backed by real artifacts.

  • PRD
  • Research
  • Post
  • Readme
  • Architecture
  • Evaluation
  • Live data

Evidence behind the Nuptis → Velora case study

Private working documents are listed with what they support; only public sources link out.

  1. Velora PRD

    PRD10 Aug 2026

    Supports: The problem, the two sides and authored Trust Scores

  2. Apparel Discovery PRD

    Research12 Aug 2026

    Supports: H1 (coordination, not effort) and the confidence tags

  3. Case Study 3 team PRD

    Research

    Supports: Pooled procurement interviews; the queue-time insight and its unverified day counts

  4. Nine-day series

    Post

    Supports: “Weddings were blue — but a shallow pool.”; the day-seven kill

  5. Nuptis PRD

    PRD7 Aug 2026

    Supports: The first bet and its unmeasured success metrics

  6. Velora README

    Readme

    Supports: The one-day build and the stack

  7. Supabase notes

    Architecture

    Supports: The live path was built but not run against a real project

  8. Final review (task 6.3)

    Evaluation

    Supports: 10/10 tests, 0 overflow, 156 kB gzip bundle

  9. Velora live app (opens in a new tab)

    Live data15 Sep 2026

    Supports: Live on mock data

Links and next steps

Solo buildAug 2026Live · mock data