Skip to content
Customer stories
PilulkaSupplier analytics

Pilulka cut supplier analytics work by 70–80%.

Customer-approved result

70–80%

reduction in manual work

95%

mismatches caught before CRM write-back

€14K

annual manual-cost saving

Pilulka's supplier analytics cycle for ~320 suppliers was mostly manual.

Before Duvo, category managers spent hours moving data between VMS, CRM, BigQuery, XLSX files, Many ERP, PDFs, and email.

Errors were often caught late, and missing approvals or evidence could stall without clear escalation.

Category managers prepared for supplier meetings without consistent data packs, supplier terms lived in emails and personal spreadsheets, and monthly evaluation/factoring work depended heavily on one analyst.

Before

  • Category managers moved data manually between VMS, CRM, BigQuery, and ERPs
  • Errors caught late and missing approvals stalled without clear escalation
  • Supplier terms lived in emails and personal spreadsheets
  • Monthly evaluation depended heavily on one analyst

After Duvo

  • 70–80% reduction in manual work across the supplier analytics cycle
  • Per-supplier briefings for ≥90% of the portfolio
  • 95% of CZ/SK and price mismatches caught before CRM write-back
  • Exceptions flagged for humans automatically

Duvo gives our teams supplier-ready briefings and catches exceptions before they become disputes, so people can focus on negotiations instead of manual data handling.

Filip Dušek

Category Manager, Pilulka

The workflow from signal to outcome.

01

Prepares per-supplier briefings

The agent prepares per-supplier briefings from BigQuery for ≥90% of the portfolio.

02

Validates consistency

Validates CZ/SK and price consistency before CRM write-back, catching 95% of mismatches.

03

Extracts printed terms

Extracts printed terms with OCR from PDFs and emails.

04

Automates the cycle

Automates guided VMS condition collection, monthly evaluation, Many ERP invoicing support, and quarterly evidence collection.

05

Flags exceptions

Flags exceptions for humans instead of asking the team to process everything manually.

Evidence status

Customer-approved result

Pilulka approved the 70–80% manual-work reduction, 95% pre-write-back mismatch detection, and €14K annual saving described in this story.

Systems involved

BigQuery, VMS, CRM, Many ERP, Excel

Control points

  • CZ/SK and price consistency are validated before CRM write-back.
  • Exceptions route to people instead of being written through automatically.
  • A 14-day parallel run validates the automated cycle.

More customer stories.

See all results