How Deutsche Telekom Uses Celonis Process Mining to Save €66M in Procure-to-Pay

Deutsche Telekom Services Europe (DTSE), the shared services arm managing over 2 million order items and 9 million invoices annually, deployed the Celonis Process Intelligence Platform across its Procure-to-Pay operations. The platform achieved €66M in total savings including €40M annual savings through a 96% cash discount capture rate, prevented €3M in duplicate payments, and extended to 1,500 employees across 150 use cases.

Impact

€66M

Total savings achieved

€40M

Annual savings from cash discount capture

90%+

On-time payment rate

€12M

Savings from automation

€3M

Duplicate payments prevented

€1.6M

Contractual penalties recovered (2019)

90%+

PO accuracy on first submission

1,500 employees, 150 use cases

Platform adoption

Challenge

Deutsche Telekom Services Europe processed over 2M order items and 9M invoices annually across 10+ fragmented source systems with no real-time visibility into cash discount capture rates, payment timing, duplicate payments, or PO accuracy — creating systematic financial leakage across Procure-to-Pay operations.

Solution

DTSE deployed the Celonis Process Intelligence Platform integrated with SAP, SAP Ariba, Salesforce, and 10+ source systems to create a real-time digital twin of its P2P processes, building a Center of Excellence for Process Mining that scaled intelligent automation and process optimization across 1,500 employees and 150 use cases.

Tools & Technologies

What Leaders Say

The vision I have is that in five years' time, nobody will need to touch our transactional systems unless Celonis suggests it.

Peter Tasev, Senior VP Procure-to-Pay, Deutsche Telekom Services Europe

Celonis is a game changer. By using process mining, we established a fully data-driven organization.

Christian Unterbusch, VP Operational and Strategic Steering, Deutsche Telekom Services Europe
Get the full context.

Sign up to read complete case studies, access detailed metrics, and unlock all use cases.

Full Story

Deutsche Telekom Services Europe (DTSE) operates as the shared services center for Deutsche Telekom Group, processing over 2 million order items worth more than €7 billion annually and handling 9 million invoices per year. At that scale, even small inefficiencies in Procure-to-Pay processes compound into material financial losses — missed cash discounts, duplicate vendor payments, blocked invoices, and delivery accuracy issues that trigger contractual penalties.

Before Celonis, DTSE lacked real-time visibility into how its P2P processes were actually executing across SAP, SAP Ariba, Salesforce, and 10+ other source systems. Teams spent time on low-value transactional work rather than exception management. The organization couldn't systematically measure what percentage of invoices were being paid on time, what share of available cash discounts was being captured, or where duplicate payments were slipping through.

DTSE implemented the Celonis Process Intelligence Platform integrated with its full system landscape to create an end-to-end digital twin of its P2P operations. Process mining surfaces exactly where payment delays, compliance deviations, and accuracy failures are occurring. The platform powers a Center of Excellence for Process Mining that scales use cases across the organization — reaching 1,500 employees across 150 distinct use cases.

The financial results are among the most documented in enterprise process mining. €66M total savings achieved to date. Annual savings of €40M driven by a 96% cash discount capture rate — up from whatever the baseline was before Celonis made discount deadlines visible and actionable. On-time payment rate reached 90%+. Intelligent automation generated €12M in savings. Duplicate payment detection prevented €3M in losses. In 2019 alone, €1.6M in contractual penalties were recovered. Purchase order accuracy on first submission exceeded 90%.

Senior VP Peter Tasev articulated the long-term vision: a state where Celonis is the trigger for every system interaction, shifting from reactive transaction processing to algorithmic process orchestration.

Similar Cases

VU
Vodafone UK
50%
reduction in copywriting time

Vodafone UK, one of Britain’s largest mobile operators, and VOIS, Vodafone’s shared services arm, deployed Writer’s enterprise AI platform to transform their marketing operations at scale. The implementation replaced manual content production workflows with AI agents that handle banner creation, campaign copywriting, and generative engine optimization. Within six months, Vodafone UK eliminated 20 hours per week of manual work and saw 2x engagement from AI-optimized campaigns, while VOIS cut copywriting time by half.

TelecommunicationsWWriter
V
Vodafone

Vodafone, a telecommunications company serving over 340 million customers, built two AI assistants using LangChain and LangGraph to support engineers managing data centers across Europe. The Insight Engine converts natural language queries into SQL for real-time performance analysis, while Enigma retrieves technical documentation from SharePoint. Together they have reduced time-to-insight for infrastructure issues and eliminated engineers’ dependence on custom dashboards.

TelecommunicationsLLangChainLLangGraph
A
AT&T
71
genai_solutions_deployed

AT&T built a multi-agent AI platform on Microsoft Azure to handle millions of daily interactions across customer care and internal workflows. The system processes roughly 9 billion tokens per day, reduced customer care resolution time by 33%, and serves over 100,000 employees across 71 generative AI solutions.

TelecommunicationsMAMicrosoft AzureOOpenAI
U
Uniti
10,000+
monthly consumer billing queries handled by ai chatbot

Uniti is a premier fiber infrastructure provider delivering mission-critical broadband connectivity across the United States. The company faced more than 10,000 monthly consumer billing queries and needed a way to give its field sales teams instant, accurate insights. By deploying Oracle Autonomous AI Database and OCI Generative AI, Uniti launched a 24/7 AI chatbot that resolves customer questions in real time—cutting support costs and improving its Net Promoter Score.

TelecommunicationsOAOracle AI Vector SearchOAOracle Autonomous AI Database
C
Cisco
1,500+
engineering hours saved monthly

Cisco embedded OpenAI Codex into production engineering workflows, achieving 20% faster builds, 10-15x defect resolution throughput, and saving 1,500+ engineering hours per month.

TelecommunicationsOCOpenAI Codex
SC
SoftBank Corp.
4,500
fte-equivalents saved annually

SoftBank Corp. is Japan’s largest telecommunications operator and a global technology investor with ambitions far beyond carrier services. Facing a bottleneck in scaling automation beyond early RPA deployments, SoftBank launched the ‘Digital Worker 4,000 Project’—a top-management-led initiative to automate cross-functional business processes using Automation Anywhere alongside Salesforce, Office365, and Google integrations. The effort redirected 4,500 FTE-equivalents of annual labor toward value-added work and generated 52,000 generative AI improvement proposals from employees in just ten days.

TelecommunicationsAAAutomation Anywhere
DT
Deutsche Telekom
~5 hours
hours reclaimed per attorney weekly

Deutsche Telekom deployed Harvey across its Legal, Compliance, and Data Protection departments, enabling attorneys to reclaim approximately 5 hours per week through AI-powered litigation support, regulatory research, and tender generation.

TelecommunicationsHHarvey
D
DSM
Multi-million-dollar
savings from pilot

DSM, the Dutch multinational in health, bioscience, and nutrition, deployed the Celonis Execution Management System to eliminate manual interventions in its Order-to-Cash process. Within a 3-month pilot, the company achieved multi-million-dollar savings, reduced delivery blocks by 10%, improved its APAC first-time-right ratio from 18% to 42%, and grew North America OTIF from 84% to 90%.

ManufacturingCPCelonis Process Intelligence Platform