Building the ultimate AI team

What are the ideal roles and skill-sets you need for an AI project? 

First of all, there needs to be an internal AI champion. There must be significant human and financial resources assigned to the project and clear leadership. Developing Artificial Intelligence is an ongoing initiative, not a fixed-term project. The executive sponsor or decision-maker of the project is a key role that helps articulate what the business is trying to accomplish and why.

The AI Champion will not usually have time to manage the day-to-day activities. This is why there should be a project owner, who will keep the focus on the KPIs and workstreams. They will also be helpful in dealing with the organisational bottlenecks and politics. 

Then there should be one or more subject matter experts. These are people within the organization who have experience with the area of the business that the AI will be utilised. They will be critical for refining the model and coming up with the parameters.

Next, there are a variety of roles on the technical side of the project. Here’s a look:

  • Data Engineer or Data Annotation Specialist: This person often gets overlooked—but this is a big mistake. Data for AI projects is usually messy. So there is a need for data wrangling and labelling, which can be tedious and time-consuming. 

  • Data Scientist or AI Engineer: They will spend time on both the data and the algorithms, such as with machine learning, deep learning and Natural Language Processing. 

  • Machine Learning Engineer or ML/Ops:  This person will spend much of his or her time productising and maintaining the models once they are deployed. Models will drift as new data comes in, so this person must understand how the models work and the data interface to the models.

  • UX/Visualization Engineer: A key reason AI projects fail is that the application is too complicated. Keep in mind that the end-user is often a non-technical user.  Thus, a UX/Visualisation engineer can make the AI much more usable. 

  • AI Testing or QA (Quality Assurance):  AI can easily break.  But an AI Testing person can help validate the model under different conditions.

  • Enterprise or Solutions Architect: This person will help with the implementation and integration of the AI project. 

If anything, the roles of the data scientist or the ML engineer are perhaps the first to focus on. They will be essential for the ultimate success of an AI model. 

Recruiting from a pool of candidates with varying technical backgrounds can create an AI team with a wide, diverse set of perspectives. This technical diversity also makes collaboration more interesting and encourages the team to effectively communicate.

It’s a great idea to look internally for candidates interested in upskilling, especially if they are Subject Matter Experts. While the talent pool is growing, it isn’t keeping up with the pace of demand. To meet the increasing need for AI talent, look internally for those with the right mindset and drive, and train them to fill those needs.

Justin Flitter

Founder of NewZealand.AI.

http://unrivaled.co.nz
Previous
Previous

Why you need to understand the three-segmented Zipf’s Law if you’re working on NLP #AI.

Next
Next

Robots that can think 10,000x faster than humans.