Selected projects for manufacturing, automotive, utilities and the public sector, each with a concrete business problem and a measurable result.
Enterprise reporting architecture
Data Analytics & Reporting
Challenge: Internal systems and external sources such as suppliers and social media delivered inconsistent figures. Management needed one reliable data basis for reporting.
Solution: A layered reporting architecture: a harmonisation layer feeds a central data warehouse, which serves a Power BI presentation layer. A data catalogue with KPI definitions and harmonisation rules governs data quality.
How it works:
- Sources
Internal systems (1–n) and external sources (suppliers, social media)
- Harmonisation layer
Mapping tables translate source codes into common definitions
- Data warehouse
Central, consistent store for all reporting data
- Presentation layer
Power BI reports and dashboards
- Governance
Data catalogue as 'rule book': data assets, KPI definitions, harmonisation rules, data mapping table
Results:
Data is correct and complete
Data is available when decisions are made
Data is consistent across all reports
Technologies: Data warehouse, Power BI, data catalogue
Faster data delivery with AI-assisted engineering
AI Solutions & Enablement
Challenge: Frequent changes to data pipelines, transformation rules and business logic can slow delivery. Understanding inherited code, resolving failures and keeping documentation current also consume valuable engineering time.
Our approach: Analyse existing notebooks, SQL and transformation logic with AI assistance; accelerate repetitive workflow and ETL changes; support debugging, root-cause analysis and documentation. Engineers remain responsible for validation, testing and release.
Principle: AI supports delivery; engineers retain accountability for quality and release.
How it works:
Change request
AI analysis of notebooks, SQL and dependencies
ETL and business-logic changes drafted with AI
Engineering validation and testing
Release
Results:
- 70 % faster workflow updates
60 Databricks workflow changes: 5 days to 1.5 days
- 50 % faster documentation
technical documentation: 4 hours to 2 hours
- 50 % faster logic analysis
business-logic backtracking: 3 days to 1.5 days
Technologies: Databricks, Databricks Genie AI, SQL, Python, GitHub Copilot
AI-assisted Power BI development with GitHub Copilot
AI Solutions & Enablement
Challenge: Business reporting evolves continuously. Changes to semantic models, DAX measures, report pages, filters and documentation can make each iteration slow and increase the risk of inconsistencies.
Our approach: Use GitHub Copilot with Power BI project files (PBIP) to understand report and model structure; draft and refine DAX, Power Query and semantic-model changes; assist updates to report pages, visuals, filters and documentation; validate calculations, data, usability and access before release.
Principle: GitHub Copilot accelerates authoring; Power BI specialists retain accountability for design, validation and release.
How it works:
Business need
GitHub Copilot on Power BI project files
Model and DAX changes
Report and UX updates
Developer validation and release
Results:
- Faster iteration
shorter path from a change request to a review-ready report
- Consistent delivery
reusable standards across measures, visuals and documentation
- Controlled release
human review of data, logic, security and user experience
Technologies: Power BI (PBIP project files), DAX, Power Query, GitHub Copilot, Git
Supplier Hub: a working solution in three days
Application Development
Challenge: Supplier information was spread across sources and handled through manual steps. Business users needed a clear, practical way to access the data and validate the process quickly.
What we delivered: We converted the requirement into a focused use case and user flow, created an AI-assisted web-application mock-up in one day, connected data and implemented the core business logic, and validated the solution with stakeholders: a working solution in three days.
Features: Supplier overview with open requests, onboarding, active suppliers and items needing attention; supplier lifecycle from request to active supplier; supplier directory, contracts and catalogues.
How it works:
Business need
1-day clickable mock-up
Data connection and business logic
Stakeholder validation
Working solution
Results:
- 1 day
interactive mock-up for early stakeholder feedback
- 3 days
working solution with user interface, connected data and core business logic
- AI + human
AI accelerated the build; specialists validated the solution
Technologies: AI-assisted development, web application
GenAI knowledge assistant for a manufacturing company
AI Solutions & Enablement
Client & requirement: A manufacturing company that uses AI-powered platforms for knowledge management, collaboration and automation. Goal: secure access to enterprise data, automated workflows via Microsoft 365 and data-driven decisions.
Challenges: Fragmented data across systems led to slow, inefficient access; traceability was limited with no clear sources; security constraints restricted generic AI tools; multiple tools caused loss of context and productivity.
Solution: GenAI-powered knowledge search in natural language across enterprise data; one conversational interface to interact with Microsoft 365; enterprise-grade security for compliant AI usage; built-in citations for transparency and trust.
How it works:
- Users
employees ask questions in natural language
- Chat interface
one conversational front end
- Backend & context
retrieves relevant enterprise content (RAG) and prepares context
- Enterprise data · Microsoft 365
documents, sites and data the user is permitted to see
- LLM layer · Azure OpenAI
generates the answer with citations
- Access control
Azure tenant, permissions and audit trail around the whole flow
Results:
- 65–75 %
efficiency gain: faster document search and response time
- 40 %
productivity: less manual effort, faster execution
- 100 %
secure: audit-ready decision support
Technologies: Azure OpenAI, Microsoft 365, Azure, Python, FastAPI, RAG
Management dashboard for operations
Data Analytics & Reporting
Purpose: An executive snapshot of quarterly performance across functional units: shipments, backlog and inventory in one view.
What it answers: How much shipped this quarter and what is still outstanding; what revenue is at risk and what has been delivered; delivery performance against plan (scheduled date versus customer request date).
Leadership value: Links operational volume, revenue impact and inventory position, so execution gaps are visible and prioritisation decisions are made faster.
How it works:
- Filters
product type, warehouse, function unit, product model, channel, customer, period
- Shipment & demand
QTD projection vs. actual, backlog, demand
- Service level
customer request date met, standard lead time met, against a 90 % target
- Business performance
booked, revenue and backlog, quarter and year actuals
- Inventory
raw material, work in progress, finished goods, in transit
- Currency
switch between € and $
Results:
Shipments, backlog, service level, business performance and inventory on one page
Execution gaps visible at a glance
Faster prioritisation in quarterly reviews
Technologies: Power BI
Customer feedback analytics on Microsoft Power Platform
Process Consulting & Automation
Challenge: Customer feedback was collected manually across several channels and stored in inconsistent formats such as Excel. Analysis was slow, incomplete and hard to act on.
Solution: A Power Platform based system that centralises feedback from all channels, structures it in model-driven applications and reports it in Power BI, enabling data-driven decisions and continuous product and service improvement.
How it works:
Channels: WhatsApp, social media, e-mail, SMS and phone
Automated aggregation with Power Automate
Structured feedback records in model-driven apps (Dynamics 365 / Power Apps)
Analytics and reporting in Power BI
Results:
- ~30 h per week
saved by automated aggregation across all channels
- One view
of customer insights for the best possible service
- Faster response
higher service quality for emerging customer needs
Technologies: Microsoft Dynamics 365, Power Apps, Power Automate, Power BI
AquaPredict: 14-day water demand forecasting for municipalities
Demand Forecasting & Planning
Purpose: AquaPredict uses historical water demand and weather forecasts to predict municipal water demand 14 days in advance.
Scope: Historical daily consumption data and weather forecasts are processed automatically in the forecasting model. Results are shown in the dashboard and sent by e-mail alerts.
How it works:
- Historical demand data
daily water demand used for model training
- Weather forecast integration
temperature and forecast data integrated automatically
- 14-day demand forecast
forecast generated 14 days in advance
- Threshold monitoring
forecast compared with defined demand limits
- E-mail alerts
automatic forecast reports and early warnings
- Dashboard & analytics
forecast, trends and alerts in one view
Results:
14-day visibility of expected water demand
Early warning for higher or critical demand
Better planning with less manual monitoring
Technologies: Python, machine-learning forecasting model, weather data API, dashboard, e-mail alerting
Digital Product Passport platform
Digital Product Passport
Purpose: We help companies create and manage Digital Product Passports to meet EU regulations (ESPR) and to drive a circular economy.
Challenge: New EU regulations require product data transparency; product information is fragmented across systems; data collection and passport creation are manual and time-consuming; lifecycle, impact and sustainability are hard to trace.
How it works:
- Data sources
ERP/PLM, suppliers, production, certificates
- 1. Data integration
collect and validate data
- 2. Data management
centralise and standardise
- 3. Digital passport
structured DPP with all required attributes
- 4. QR code
unique code for access and sharing
- 5. Analytics
dashboards and reports
- Output
DPP portal, QR access, customer view, compliance, sustainability
Results:
EU compliance with ESPR and future regulations
One source of truth for product data across the lifecycle
End-to-end traceability
Support for reuse, repair and recycling
Faster to market through automated processes
Technologies: ERP / PLM integration, QR codes, Microsoft Azure, Power BI
Automated purchase order processing in SAP
Process Consulting & Automation
Challenge: Purchase orders arrived by e-mail and as Excel files and were typed into SAP by hand. This took time and caused entry errors.
Solution: An automated workflow reads incoming orders from e-mails and Excel attachments, checks the data and creates the orders in SAP. Exceptions are routed to a clerk for review.
How it works:
Monitor the order mailbox and extract attachments
Read order data and map it to SAP fields (customer, material, quantity, delivery date)
Validate against master data and business rules
Create the order in SAP via interface or RPA
Exception queue, logging and status reporting
Results:
- 70 %
less manual entry time
- 95 %
improvement in data accuracy
Technologies: SAP, RPA, Excel
Crash simulation for battery housings
Engineering: design validation & simulation
Challenge: Battery housings must protect the cells in a collision. Physical crash tests are expensive and come late in development.
Solution: Virtual crash simulations identify weak points early. Geometry and material are improved in several iterations before physical prototypes are built.
How it works:
CAD preparation and meshing (HyperMesh)
Material models, boundary conditions and crash load cases according to requirements
Crash simulation
Evaluation of intrusion, deformation and stresses
Design iterations and recommendation report
Results:
Maximum safety and protection in a collision
Shorter development time
Lower development cost through fewer physical prototypes
Technologies: HyperMesh, CATIA
FAQ chatbot with 24/7 answers
AI Solutions & Enablement
Challenge: The support team spent much of its time on the same recurring questions, and customers had to wait outside office hours.
Solution: An AI-powered chatbot answers frequently asked questions instantly and around the clock, and hands over to a person when needed.
How it works:
Knowledge base built from FAQs and documents
Language model with retrieval on approved content
Website integration
Handover to a human for complex requests
Analysis of unanswered questions to improve content
Results:
24/7 support
Higher customer satisfaction
Relief for the support team
Online dashboard for EU university programmes
Data Analytics & Reporting
Challenge: Data on study programmes across EU universities was complex and difficult to compare.
Solution: An interactive online dashboard visualises the data and lets users filter and compare programmes to gain insights.
How it works:
Data collection and consolidation
Common data model for programmes, institutions and countries
Interactive web dashboard with filters
Publishing for online access
Results:
Complex data visualised clearly
Valuable insights to optimise study programmes
Automated Power BI reporting for dealer performance
Data Analytics & Reporting
Challenge: Dealer reports were extracted and distributed manually, which was slow and delayed decisions.
Solution: Report creation and distribution are automated with Power BI. Each dealer receives current, data-driven insights into performance and efficiency.
How it works:
Automated data extraction from source systems
Power BI dataset with dealer KPIs
Scheduled refresh
Row-level security so each dealer sees only own data
Automatic distribution via subscriptions or exports
Results:
No manual report preparation
Significantly better dealer performance and efficiency
Technologies: Power BI
Smart field service scheduler
Demand Forecasting & Planning
Challenge: Field service planning was done manually, leading to long routes, uneven workloads and high operating costs.
Solution: A scheduler uses real-time analytics to plan routes, assign technicians and allocate resources.
How it works:
Job, technician skill and location data
Optimisation of assignments and routes
Real-time updates when jobs change
Dashboard for dispatchers
Results:
Lower operating costs
Higher service efficiency
Optimised routes, technician assignments and resource allocation