The business intelligence (BI) landscape is changing. Traditionally, BI analysts used dashboarding and analytics tools like Microsoft Excel, Microsoft Power BI and Tableau. However, analytics tools are rapidly evolving, and BI analysts are expected to evolve alongside industry advancements. In particular, predictive analytics used to be in the domain of more technical employees, but today, no-code automated machine learning (AutoML) tools mean anyone can deploy AI.
Looking at BI Analyst job offerings, we can see that many now ask for AI skills because they want someone who can deploy predictive models that actively impact the company, instead of passive analytics like dashboards.
For example, several LinkedIn job postings for “business analysts” and “business intelligence specialists” now ask for knowledge in machine learning, modeling and predictive analytics.
What Are The AutoML Tools To Learn?
There’s a huge variety of AutoML tools out there, which begs the question: What should you learn? Fortunately, the best AutoML tools are the easiest to figure out, so these are nothing like learning Python, R or SQL.
If you’re less technical, then no-code tools like Obviously.AI are the way to go because you can start making predictions from tabular data in minutes, with virtually no set-up process.
RunwayML is my go-to for creative, video or image-based applications of AI, such as generating images and recognizing objects.
The Verge described GPT-3 as an “auto-generating text program,” and the GPT-3 Playground offers simple GUI-based input-output functionality, so it’s not a stretch to call GPT-3 an AutoML tool as well.
If you’re more technical, then Google’s AutoML suite may be more suited to your needs because it’s extremely feature-rich. That being said, it will naturally take a lot longer to get up and running, as well as to maintain. You’ll also want to be careful not to rack up expenses because pricing isn’t as straightforward as some of the other options.
The biggest mainstream competitors to Google’s AutoML include Microsoft’s Azure AI and Amazon’s AutoGluon, which are similarly feature-rich, but complex.
How To Get Started
Deploying AI in your organization is first about creating clarity of purpose. What business challenges are you facing? What data do you have available? If there’s an overlap between your answers (e.g., if you have data around KPIs like customer churn, traffic or conversions), then that may be a ripe use case to tackle.
Most likely, your organization analyzes many KPIs, and you’ll have tabular data around several. In that case, you’ll want to prioritize the KPIs that are most important to your organization, and score use cases accordingly. For instance, traffic may be less important to you than conversions, so you’d want to tackle the latter use case first.
Anyone in a technology-related position, including in business intelligence, should be a lifelong learner and make sure they’re staying up to date with tools that can help their organizations succeed.
Today, that means learning AutoML tools, finding meaningful use cases and deploying AI.