Unlock value from your financial data by using AI and the Cloud

Defining and implementing a Single Point of Truth for our client in the food and beverage industry

100%

automated workflow execution
75%

time reduction for fixing master data issues
95%

accuracy of product and customer models
90%

of algorithms can be reused for other markets

Arash Besadi
Data Transformation & Analytics Manager
PwC Switzerland

Giridhar Joshi
Data Transformation & Analytics Director
PwC Switzerland

The finance team of a global Food & Beverage company was looking for a solution to harmonize their master and transactional data. Their data was stored on several Enterprise Resource Planning (ERP) and Enterprise Data Warehouse (EDW) systems across various markets due to their highly decentralized IT landscape.

As a result, the finance team was forced to spend hours performing manual mapping exercises, which were tedious, time consuming and error-prone. That’s why we created a single point of truth (SPOT) platform that was tailored to the client’s specific environment and business needs.

The solution is built on an end-to-end architecture using both artificial intelligence (AI) and cloud functionality to provide automation and scalability simultaneously. In addition, the harmonised financial data is displayed on an interactive business information dashboard and is easily accessible to stakeholders.

As a result, the finance department has a better overview of its financial data, which has sped up corporate decision-making and improved the accuracy of monthly forecasts.

Key challenges for our client

The four main problems the Finance team faced when completing their monthly reporting exercise were as follows:

High level of manual coordination required, long waiting times for processing and reporting

Fragmented data source system landscape (globally)

Decentralised and non-standardised data model

Unclear data governance and lack of data ownership

The solution we created together with our client

To address these points, we built an intelligent and scalable SPOT:

AI and Cloud driven transformation

By using a three-layered (Bronze, Silver, and Gold) data lake approach, the raw, unstructured data is transformed and aggregated into a well-constructed dataset ready for the Finance team to use for further analysis and downstream processes.

The data harmonisation is supported by three sub-processes:

Natural Language Processing (NLP) and ML algorithms to classify the local entities into their corresponding global entities by using semantic similarity techniques like cosine similarity.
Cognitive search based on various AI capabilities that use web search engines and external databases to augment the internal data.
A user feedback loop to teach the algorithms which entities are false positive classifications to increase the accuracy of the platform over time.

For reporting, the users are provided with an interactive dashboard where they can track the current harmonisation progress as well as various other financial KPIs.

Benefits for our client

By using our SPOT platform, we enabled the client to decrease their costs, increase speed, and reduce manual workload.

Improved reporting accuracy

Improved accuracy on budgeting and forecasting processes through standardisation and population of data attributes across product and customer data

Centralised data

Reduced the effort required to maintain data in several ERP and SAP systems by creating a SPOT. This can be further utilised for reporting and analytical purposes