Getting the AI & ANALYTICS team right

For companies that weren’t born in the digital age transforming into a data-driven organization is a hell of a job. It requires the right technology, the right data, the right agile culture, the right business vision, the right transformation, the right AI methodology and the right AI & Analytics team. History confirms the complexity of this kind of challenge. It took about 30 years to move from steam driven factories to modern day electric engine factories. Most steam engine based companies did not survive this transformation. Let’s eat the elephant in pieces. In this blog we zoom in on one aspect, getting the AI & Analytics team right. 

Your AI & Analytics dream team

The AI & Analytics domain is evolving quickly. It now encompasses technical topics like extreme data, building data pipelines, creating algorithms, training AI models and  contextual understanding. Business topics are essential, like improving performance with specific use cases, building your data-driven organization, creating agility and embedding AI & Analytics in all of your business processes. On top of this we find social topics like compliancy, fairness and ethics have become condition sine qua non in this field.  

Inevitably therefore the right AI & analytics team is a multidisciplinary team. Yes, even a kind of multiskilled dream team in our vision. A team consisting of very different, complementary roles & capabilities, with open minded people.  Sounds terribly logical but what does this team actually look like and how to make it thrive?

Key roles

Prime objective of any AI & Analytics team is to create business value through predictions and insights. As technological possibilities are relatively new to mankind, no surprise that many organizations still need to find out what the best possible business outcome exactly looks like. And how to create this best. This is where subject matter experts, business managers, AI business consultants, AI experts and data engineers need to meet in a creative process. 

This process is typically driven by the AI business consultant, sometimes known as the analytics translator, making the different domains collaborate. A design thinking based approach delivers the best results in our experience. Combining AI technology, available data with products and services,  business processes, business models and human resources. The next step is to validate the concept and create its first value by building and shipping a Minimal Viable Product.

The data product needs to be fair, ethical and compliant with internal procedures and legislation like GDPR. This  is where the data privacy and security expert comes in. Advising and monitoring the team, and working with the data privacy officer, the security officer and the ethical board.

For the largest part of their time an AI & Analytics team is preparing, cleaning and analyzing the dataset to create the data product. The more mature organizations have implemented supporting tools (like e.g. Google’s AI hub, Data catalog and Kubeflow), organized their data and data governance, and made it as easy as possible to find and access data in a controlled manner. All this greatly reduces the time spent on preparational tasks.  

AI & Analytics team roles

• • • • • • • • AI business consultant
              • Data scientists
              • AI experts
              • Data engineers
              • Data privacy & security expert
              • Solution architect
              • SW programmer
              • UX designer
              • Success manager

Still in the AI & Analytics team it is advised to have specialists on board to execute these tasks, the data engineers. They excel at transforming, shaping, filtering, moving and connecting the data. Leaving space for the complex and unique statistical methods to the data scientists and the AI model training to the AI experts. 

Making the data product available at scale to targeted end-users and/or customers requires yet another set of capabilities in the team. On the technical side we need a solution architect to ensure the solution fits with the organization’s  IT landscape and delivers the intended performance. A SW programmer is needed to integrate the data product with the input data and to integrate the outcome with the applications used by the end-users and/or customers. To deliver a dataproduct that meets today’s ease of use standards a UX-designer is onboarded.  

Managing the success of your data product by clearly communicating with stakeholders and if needed providing training to stakeholders is an essential task. This is the task of the AI business consultant or a specialized success manager. 

How to make your AI & Analytics team thrive?

The right environment is needed to make your AI & Analytics team thrive. Sponsoring, mandate, subject matter expertise, business support, interpretation of findings, IT support, changed business decisions as a result of insights or predictions are key items typically situated outside of the team. 

Successful organizations create the right environment. In all the right environments you will find the following.  At the highest level a Board that provides the AI & Analytics function with guidance, support and mandate. On executional level a product owner who prioritizes to get the most important features of the data product right. Subject matter expertise is available to inject specific knowledge. The IT department supports the AI & Analytics team by providing data, tooling, IT support and integration into the existing IT landscape. You have implemented the agile way-of-working as the default standard. An innovation friendly culture, digitally skilled employees and most decision making low in the organization?…. check!   

All of the above enables your AI & Analytics team to really thrive. They are also the building blocks which you will find in any data-driven organization.