Skip to content

Demo portfolio

Black-and-white portrait of the founder in a turtleneck

Move your cursor to suit up the rider

Blue Rider · Portfolio

Demo portfolio

We show our method in two independent demonstrations. Neither is client work, and there are no testimonials or commercial results to present.

A1DEMO

B2B research from public sources

Independent demonstration — not client work.

A sample of 20 accounting firms in southern and southeastern Brazil, researched on their official websites on 10 Nov 2026. Below is the format only, with names omitted.

12 verified
5 partial
3 unverified

Criteria: Inclusion

  • Active official website describing the service on its own page.
  • Visible city.
  • Public institutional channel.

Criteria: Exclusion

  • Directories and shared portals.
  • Social-media-only profiles.
  • Duplicates by domain.
  • Personal-only contact.
B2B research from public sources — Independent demonstration — not client work.
#SegmentCity/StateInstitutional channelCheckedStatusGaps
1AccountingPorto Alegre/RSContact form10 Nov 2026VerifiedSize not verified
2AccountingPresidente Prudente/SPContact form10 Nov 2026VerifiedSize not verified
3AccountingBelo Horizonte/MGContact page10 Nov 2026VerifiedSize not verified
4AccountingFlorianópolis/SCContact form10 Nov 2026VerifiedSize not verified
5AccountingCuritiba (state not read)Contact link10 Nov 2026PartialState not read on the page
6AccountingNot confirmedNot confirmed10 Nov 2026UnverifiedServer blocked access; we did not bypass it

Company size was not verified for any record. The full sample, with names and URLs, can be shared in a conversation.

A3DEMO

Data cleanup

Entirely fictional demonstration — not client work.

An invented 15-row table with duplicates, stray spaces, inconsistent dates, empty fields and badly formatted values. See the before and after.

Before

Data cleanup: Before
IDNameDateValue
C002› Bruno Lima2026-03-18980,00
C004Diego Alves22 de março de 2026R$ 75,9
C005Ana Souza (exact copy of C001)15/03/2026R$ 1.250,50
C003Carla Dias03/04/20261250.5
C006Elisa Prado30/03/2026mil e duzentos
C009›Gabi Nunes‹(empty)1,250.50

After

Data cleanup: After
IDNameDateValue
C002Bruno Lima2026-03-18980.00
C004Diego Alves2026-03-2275.90
C005removed (exact duplicate; kept in the log)——
C003Carla Dias03/04/2026 ⚠ ambiguous1250.50
C006Elisa Prado2026-03-30mil e duzentos ⚠ text
C009Gabi Nunes(empty, kept) ⚠1250.50

›‹ marks stray spaces. ⚠ marks a decision left to the client: nothing is guessed or filled in.

Rules applied

  1. R1Trim and collapse spaces.
  2. R2E-mails in lowercase.
  3. R3An unambiguous date becomes YYYY-MM-DD; an ambiguous or invalid one goes to human review.
  4. R4A recognizable value becomes 0.00; an ambiguous or text one goes to human review.
  5. R5An exact duplicate leaves the clean version and stays in the log.
  6. R6An empty field stays empty.

Quality panel for the A3 demonstration

DEMO · fictional data

Indicators from the fictional demonstration. They are not real company metrics.

28

Automatic corrections logged

3

Exact duplicates removed

10

Items for human decision

  • 15 input rows = 12 kept + 3 duplicates removed.
  • No empty field was filled in.
  • Running the cleanup again changes nothing.
  • Every input ID is in the result or in the removed log.

Want to start with a small pilot?

Talk to ARDVEK on business WhatsApp, or look at the demonstrations before you reach out.

Business contact through ARDVEK’s WhatsApp. This site has no active form or e-mail address.