01

Business Analysis · Public Procurement · Data Analytics

EU Digital Procurement Competition Monitor

An audit-ready decision-support system designed to identify competition patterns and supplier concentration across EU digital procurement notices using official TED data.

Independent end-to-end case study · Public data · Documented stakeholder scenarios

60–90 second overview

The case at a glance.

Business question
How can public-procurement teams screen digital procurement results for competition patterns while retaining a traceable, reproducible evidence trail?
Scope and role
I documented requirements, defined the evidence cohort and KPI rules, and connected the analysis to data checks, UAT and reporting outputs.
Intended users
Procurement directors, category managers, supplier-management leads and data-governance reviewers.
Key result
A reproducible screen across 38,971 notices, including a 47.7% observed single-bid rate within 27,100 competition-evidence notices.
Guardrail
The screen prioritises follow-up questions; it is not evidence of fraud, non-compliance or supplier risk.

An analytical decision

Keep the denominator visible

The single-bid rate uses the 27,100 notices with competition evidence, not all 38,971 notices. Keeping that boundary visible prevents missing evidence from being treated as a competitive outcome. The result is a screening signal for further investigation, not a finding of misconduct.

Review the KPI definition, evidence cohort and validation checks.
38,971EU-27 digital result notices analysed
27,100competition-evidence notices
47.7%observed single-bid screening rate
€72.7Bquality-screened known EUR award value

Analytical chain

01Business question02Evidence03Model & KPIs04Decision support

Project evidence

Artifacts that make the work reviewable.

The dashboard is a real repository output. The traceability and UAT previews reproduce verified rows from the project documentation.

01

Problem & scope

How can public-procurement teams screen digital procurement results for competition patterns while retaining a traceable, reproducible evidence trail? EU-27 contract-result notices for digital procurement, scoped to CPV divisions 48 and 72 and published from 2025-01-01 to 2026-07-30 in the verified build. Official TED Search API result notices scoped to EU-27 digital procurement. A reproducible source manifest preserves the extraction context and public-data provenance.

02

My contribution & intended users

I documented requirements, defined the evidence cohort and KPI rules, and connected the analysis to data checks, UAT and reporting outputs. Procurement directors, category managers, supplier-management leads and data-governance reviewers.

03

Requirements & calculation rules

Business requirements define scope, reproducibility, calculation behaviour and the evidence required before a KPI can be displayed. Every reported measure has a defined numerator, denominator, cohort and data-quality condition. Known award value is kept separate from records where currency or value cannot be compared reliably.

04

Analysis & outputs

Management questions were translated into analytical cohorts, explicit KPI rules and acceptance criteria. Python and SQL pipelines apply controls before data enters the reporting layer. A SQLite analytical model separates notice, buyer, supplier and award perspectives so measures can be reproduced and interrogated without relying on one flattened export. The portfolio analysis observed a 47.7% single-bid screening rate within 27,100 competition-evidence notices and analysed €72.7B in quality-screened known EUR award value.

05

Validation & traceability

UAT scenarios connect requirements to transformation rules, SQL outputs and published measures, creating a clear route from question to reported result. Coverage checks, currency handling, deduplication rules and explicit exclusions protect the interpretation. A single-bid signal is a screening indicator—not evidence of fraud, corruption or misconduct.

06

Recommendation & limits

The output helps prioritise follow-up questions about market participation, supplier concentration and data completeness. It does not determine wrongdoing or legal compliance. The analysis depends on the completeness and consistency of published TED fields. It is a portfolio study of public records, not a regulatory investigation or claim of commercial impact. The screen prioritises follow-up questions; it is not evidence of fraud, non-compliance or supplier risk.

07

Technologies and Repository

PythonSQLSQLiteTED Search APIData Quality ControlsBusiness Analysis Documentation
Review the complete repository

Continue exploring

Next case studyE-Commerce Sales, Cancellations & Customer Value Decision System