From the inception of Kepler, the Stradigi AI team has kept a razor-sharp focus on ensuring accessibility for our platform. As we’ve refined our vision on what this actually means, the most common thread of conversation among all departments has been: how can we create something for users with no machine learning experience? In other words, how can we begin to break down the barrier of entry to ML?
At first, exploring this territory was dominated by a vision to empower Business Analysts with Machine Learning. Based off of conversations with clients and partners, it seemed simple enough: analysts have data they need to use, and are in the business of constantly generating insights to improve company performance. They’re often the folks creating key insights like revenue forecasts, efficiency-driving tactics, and sales predictions. When you think about it, it’s a context perfectly fit for an ML platform: plenty of data, a need for automation, and a direct line to business impact.
But, as the use cases that Kepler can solve continued to proliferate, a broader picture of our target user started to emerge. So we dug a little deeper.
Recently, we welcomed various types of analysts to our offices at Stradigi AI, all from around Montreal. They came to discuss a few key topics related to their lives at work. One of the key themes that emerged from this session was that the word “analyst” has hundreds of implications, which are invariably impacted by location, industry, seniority, business context such as size and department, to name a few. All of these individuals range in their technical experiences, from coding to citizen data science work, to SQL queries, or, in many cases, none of the above.
There’s one universal item that connects analysts, though: they all have a deep awareness of the business, and their specific role to understand what challenges are, and subsequently seek data-driven ways to solve these challenges. Here’s the catch: today, isn’t that true for any internal Subject Matter Expert?
We’re taking the stance that it is. And with that logic in mind, Machine Learning can be for anyone with swaths of data who wants to solve pressing business challenges that result in a real, measurable impact. Those people are already sitting in desks right beside you. Still not sure? Consider the points below:
1. No one knows your business better than your internal people.
The shortest pathway to implementing AI successfully is starting with a clear and tangible business problem. That business problem then becomes your Machine Learning use case. And who knows your business problems, better than the people working in your organization, every day? Leaders have become too distracted on the novelty of building RNN transformers, convolutional neural networks, heuristic search techniques, and so on. But if it’s possible to automate the complex Machine Learning and Data Science tasks, the only thing you — and your people — need to keep a razor-sharp focus on is your goals, and what’s standing in the way of you reaching them.
Plus, if you have seasoned employees who have been with you for, say, over a decade — they’ve already seen what happens to your company in a variety of conditions, from hypergrowth to managing through an economic downturn.
That level of experience is immeasurable. And your teams already know who to go to for the information they need, and they’re acutely aware of your long term vision. Kepler requires answers from internal SMEs as part of its Automated Data Science Workflow, because we know that successful Machine Learning starts with a broad, well-informed picture of your business.
2. No one knows your data better than the people who’ve created it, managed it, and laboured over it.
On Thursday, Stradigi AI’s Director of Solutions, Maria Elena, discussed the importance of shifting narratives from “data readiness” to “data ecosystems.” Not only is data highly dependent on departments, there are specific people who hold the keys. We know that building company-wide data literacy doesn’t happen overnight — but educating our teams on how to better use the data they own and care for helps create deeper understanding of the true impact that data can have on reaching goals. If you can equalize access to Machine Learning tools, you’re getting one step closer to bringing your entire team even closer to the data they use every day, and augmenting their data literacy in the process. If you alleviate time-consuming data tasks like data pre-processing and cleansing, that’s an added bonus.
3. No one knows how to communicate results, needs, and impact to stakeholders better than your current teams.
Last week, I discussed the importance of using ML to “upskill” your teams, and how empowering them with Machine Learning tools not only makes them feel like they’re a part of the process of AI transformation – but gives them the opportunity to say something more with their data. Whether it’s a Marketing Strategist speaking to a CMO about optimizing promotional offers, or an Operations Manager delivering critical insights on organizational efficiency, the business context, relationship, and historical knowledge these subject matter experts hold can’t be underestimated. Machine Learning is nothing without data, and data is useless without a solid foundation and meaningful story behind it — and your people hold those keys.
Edited by Colleen McNamara