Data Science – What decisions are you hung up on making?
Better decision-making hinges upon the data upon which it is based. Many reporting systems are similar to a flight recorder: they document past events retrospectively without further reflection. You can’t chart a path with that. We work with your team to address the complexity and dynamics in order to ensure decision-making capabilities over the long run.
- How do we arrive at comprehensible, data-driven decisions?
- How do I deal with peak workloads that bring new questions, new regulations or new technologies for my team?
- Which of the many methods and (cloud) technologies existing are the right mix for me?
- How centralised/decentralised do we want to be?
- How good is our data quality and how can we promote it?
- How do I get rid of old reports?
Business Intelligence makes it possible to analyse data efficiently so it can be presented in a usable manner and used as the basis for making well-informed business decisions. These analyses can then be taught to learn using artificial intelligence. While BI generates reproducible data paths that always deliver the same output with the same input, AI can learn from the BI data pool and thereby improve the accuracy of predictions and decisions. That’s why one topic should not be considered without the other. Successful data science projects always examine both sides and then look at the specific application to decide which technologies and tools can achieve the best possible result.
Our Data Science Services
Successful data science projects need more than just databases and algorithms. We combine our knowledge of problems with methods and experience and build bridges for you:
We demystify and identify data capabilities
- Big data and AI: what are they all about? Making methods and capabilities understandable
- ‘What’s a good opportunity?’, ‘What can I do with my data?’ – as a capabilities workshop, for example, or in a proof of concept
- Customer segmentation, forecasting, anomaly detection, logistics optimisation
- AI topics often focus on process automation, but explainable AI (XAI) allows for transparency as well as decision-making support.
We help design your data architecture
- Architecture workshop to create modern, robust and scalable BI and AI solutions, both on premises and in the cloud
- Help with implementing big data architecture and AI technologies in a responsible manner
- Modernisation and reduction of costs by migrating ETL processes from the host to UNIX/Linux-based systems
We implement Data Science on your behalf
- We assist you with your machine learning project as a data scientist, BI specialist, architect, reviewer or project manager, or implement it on your behalf. You can hand over the entire BI task to us and we’ll also ensure its efficient operation.
- Prototyping, functional specification, requirements analysis, project management and quality assurance
- Implementation of the entire pathway of data sources, ETL, DWH to machine learning, from SAS/Base to R, Python and TensorFlow and cloud deployment
We optimise technologies and methods
- Advice on how to maintain and optimise existing BI and AI systems, from a proof of concept in Jupyter Notebook to enhancing the production-ready AI pipeline performance of ETL processes
- Automated testing of the entire BI stack, approval support for machine learning models
- Tamper-proof and reliable deployment of BI data paths and data science pipelines
- Optimisation of the maintainability and expandability of all data science processes, capability maturity models
- Introduction of processes for central bug tracking and debugging
- Introduction of data vault methods
We make the development of data science agile
- How do you organise a data science group? How do we identify the controlled degrees of freedom required by data science projects?
- Ability to be quickly expanded if changes are made to the source systems’ data model
- Ad hoc evaluations by specialist departments, sandbox reporting, self-service BI
- We will gladly help you with a project structure set up according to functions: teams or individual consultants work on specific applications right from the requirements analysis to implementation. This shortens feedback cycles in projects and promotes learning from the data!
Data Science seminars for management, data scientists and data engineers
“If glasses improve perception and security, then explainable artificial intelligence (XAI) must be the ‘glasses’ for data.”
Successful data science projects with viadee
AI methods have matured. The question is often no longer about feasibility. Instead it’s about integrating AI models into your business processes and your data landscape in a sustainable manner. As a partner who’s specialised in corporate and IT architectures and in the digitisation of business processes and their automation for many years, viadee has extensive experience in these areas.
Our consultants take a broad perspective. Although it is becoming far easier to create proof of concepts, it’s still a challenge to identify business cases and fully develop them into a productive operation (ML Ops). We can help you with this.
Depending on the area of application, reliable and transparent decision-making is required by law. That’s why we provide an analysis explaining why AI models arrived at certain decisions (i.e. ‘Explainable AI’ [XAI]) for each and every project. We work with you to take a deeper look at learned models, since methods such as deep learning often make meaningful decisions.
Each decision and projection made by our AI models can be analysed on a case-by-case or general basis and the factors used as input can be explained. We clarify artificial intelligence’s knowledge and the trigger for making a decision to you so you can use this know-how about new rules and causalities independent of automatic processes.
COMPANIES FOR WHICH WE HAVE SUCCESSFULLY IMPLEMENTED DATA SCIENCE PROJECTS:
In search of inspiration?
Data science offers us a wealth of opportunities that we’d be happy to try out with you. It would be our pleasure to help you identify a specific use case. We regard data science primarily as an aid and guide into the widest possible range of application areas. That is why we work closely with the other expert viadee teams, from business process management and cloud to test automation.
If glasses improve perception, then explainable artificial intelligence (XAI) must be the ‘glasses’ for data.
SAS is a market-leading provider of statistical analyses, data mining, data warehouse solutions and business intelligence. As a long-standing SAS Silver Partner, we possess a great deal of knowledge about a wide variety of SAS products and can look back on numerous successful SAS implementations and SAS projects with our customers.
Does your business already use SAS? We’d love to use our certified know-how to provide advice on new SAS-based developments or architecture. Do you already know what the new SAS Viya can do? As a SAS partner, we were already able to try out this product. Feel free to ask us about it – without any further commitment!
If you’re contemplating using SAS, we would be happy to share our experience with you and provide you with impartial advice. We don’t deliver ready-made solutions – our top priority is to merely listen.
What is your project? Does it involve a requirements analysis, data modelling and extraction, developing an interactive BI portal or integrating SAS into big data architecture? Together we will find a good solution.
MODERNISE DATA WAREHOUSES AND PASS ON THE RESPONSIBILTY
Has your system grown over the years and is no longer flexible enough to keep up with current requirements? Does the integration of new data sources and technologies represent major challenges for your data warehouse? Are your data quality and performance no longer up to par?
A number of different approaches exist to satisfy the requirements of a modern data warehouse. Such examples include data virtualisation, data vaults, big data, data warehouse automation, and agile procedures in the data warehouse environment. We will find the right solution for you using our experience gathered from a wide variety of projects.
It would be our pleasure to take on complete responsibility for your data warehouse. In each case, a viadee team of around four to eight consultants oversees day-to-day operations as well as continued technical development. We flexibly adjust our team to your requirements.
Tell us your ideas and plans – we collaborate with you to find the solution that best fits your requirements, your technical set-up and your budget!