CASE STUDIES
CASE STUDY
Audit-as-a-Service
Client Background
In various jurisdictions, firms are required to undergo annual audits by licensed, independent audit firms. These auditors must remain neutral and have no affiliation with the companies they evaluate.
Project Overview
Challenge
Traditional audits are often performed after the fiscal year has concluded, leaving auditors with limited time to analyze the company’s data and uncover potential issues. By the time any discrepancies or fraud are detected, it is often too late to mitigate the damage. In an era of AI and advanced IT solutions, the question arises: why aren’t audits being conducted in real-time, or immediately after incidents occur?
Approach
As part of a research project at the University of Osnabrück, where Dr. Johannes Langhein worked from 2017 to 2020, we developed a cutting-edge concept called “Audit-as-a-Service.” This cloud-based platform allows audits to be conducted in real-time, enabling auditors to access crucial information on-demand. The IT infrastructure is managed by a specialized service, so auditors can focus on their review while clients receive timely updates. This proactive approach ensures that potential issues are identified early, reducing risk and minimizing damage to the company.
Results
This concept was revolutionary for the audit industry and sparked a wave of innovation among startups and established IT firms. While there is still progress to be made, traditional auditing methods—some of which are decades old—are now being challenged by data-driven, AI-powered solutions. In Germany, Audit-as-a-Service initiated a critical conversation about how auditing can evolve in the 21st century.
Testimonial
These key findings are based on a 2019 study evaluating the impact of Audit-as-a-Service:
“Audit-as-a-Service saved valuable human resources, reduced costs, and offered a superior alternative to in-house solutions. Highly recommended!”
CASE STUDY
AudITScraper
Client Background
A key responsibility of CPAs and accountants is to create accurate financial statements that reflect the financial activities of companies. These statements are later reviewed by auditors for correctness, completeness, and legality. Every business transaction must be recorded promptly, which adds complexity when managing data from multiple systems.
Project Overview
Challenge
For accountants and CPAs, one of the biggest hurdles is gathering information on business transactions from various sources. Auditors face an even tougher challenge, as companies often use different systems and accounting methods, making it difficult to validate the data. Additionally, errors such as false, incomplete, or damaging transactions can lead to significant financial harm if not detected early.
Approach
As a senior researcher at the German Research Center for Artificial Intelligence (dfki), Dr. Johannes Langhein led the development of a pioneering solution called AudITScraper. This software proved that auditing critical business transactions before they enter the ledger is not only feasible but highly effective. The system requires three random auditors to approve a transaction before it can be recorded, leveraging AI to detect potential risks. This innovative approach demonstrated a significant reduction in financial risks and the prevention of costly errors.
Results
We successfully implemented a web-based tool that uses AI to detect anomalies in business transactions from various pre-systems. The tool creates an environment where auditors can access detailed information about high-risk transactions, allowing for informed decision-making and reducing the chances of financial damage to companies.
CASE STUDY
WindeeAI
Client Background
Windreserve manages and maintains approximately 20 wind farms in Northern Germany, focusing on turbines that are 15 years or older. The company operates with three maintenance crews, consisting of 10 service technicians, electricians, and IT specialists. They also developed Windboxes, which ensure reliable and secure retrieval of SCADA data from their wind farms.
Project Overview
Challenge
Every summer, before the peak wind season, Windreserve faces a critical decision: whether to invest in their aging wind turbines or risk extended downtimes. To make informed choices, they need up-to-date information on machine condition, recent service history, repair costs, and current performance metrics. Accessing this data is challenging, as it comes from multiple sources, and some data doesn’t even exist, leading to poorly planned service trips and incorrect equipment being used.
Approach
As the project leader of WindeeAI, Dr. Johannes Langhein designed and implemented a database and data flow system to gather and store data from various sources and APIs. Leveraging this infrastructure, his team developed WindAgent, an AI-based agent that provides real-time insights on machine condition, event codes, oil status, and more to the service technicians. Additionally, the team built a modern WebConsole that allows technicians to monitor live data, generate reports, and manage service calls efficiently.
Results
Service technicians can now access critical information much faster than before, including data that was previously unavailable. This improvement has led to more efficient service planning and execution, significantly reducing time and costs. As a result, other wind farm managers have expressed interest in adopting this solution.
