No Code Required: The Rise of the Citizen Data Scientist

When a restauranteur looks to hire employees at their popular eatery, they want servers who can make a cocktail, bartenders who can bus a table, and busboys who can take an order. Why? This kind of flexibility and teamwork builds camaraderie and satisfaction, as individuals work together toward a shared goal. Plus, the restaurant is more resilient to crises like the Friday dinner rush or last-minute sick days.

For BFSI orgs today, it’s much the same. Given the recent back-to-back crises of Covid-19, the Great Resignation, war in Ukraine, and looming recession, role versatility makes companies more resilient in the face of further macroeconomic turmoil. That’s why forward-looking enterprises are arming their non-technical employees with hardcore AI and ML tools that were previously under the sole purview of data scientists. These “citizen data scientists,” in turn, provide a cascade of benefits across operations.

What is a Citizen Data Scientist?

According to Gartner, a citizen data scientist is “a person who creates or generates models that leverage predictive or prescriptive analytics, but whose primary job function is outside of the field of statistics and analytics.” In other words, these are non-technical employees who perform tasks traditionally in the domain of engineers—building, fine-tuning, and running experiments with machine learning models—without code. This is a new era of AI.

Why Deploy Citizen Data Scientists? A Cascade of Benefits:

1. Ease-of-Use

The very premise of the citizen data scientist rests on ease-of-use. Complex data workflows that once required technical expertise are now open to non-technical roles because in lieu of code, users create models with simple, user-friendly drag-and-drop interfaces. This isn’t just a “nice to have,” either. “As the requirement for digital skills expands across the workforce,” says EY in a recent Tech Horizon report, “the key criteria for new technology deployments will be ease of use and self-service by the user.” So as BFSIs continue to deploy AI, they’ll be smart to pick a platform with a no-code/low-code UI.

It’s not unlike website-builder Squarespace. Their platform revolutionized website creation by combining ease-of-use and advanced capability. Both a Fortune 500 enterprise and your average Joe can make a good-looking, professional website with the same tool—quickly and cheaply. Consider Joe, in this example, a “citizen web developer”—and AI is getting Squarespace’d.

2. Democratize AI

The benefits of automation are well-established, but there are, of course, barriers to entry. Procuring executive buy-in is one thing, but ensuring you have the talented technical personnel to manage these programs is another. A no-code/low-code approach to automation lowers these barriers, not only by opening easy-to-use automation tools to non-technical members of your team, but making automation more accessible to smaller companies that don’t already have a deep bench of ML engineers and data scientists.

3. Bridge the Talent Gap

We’re living through a historic talent crisis. And there’s a series of discrete, complex reasons for the shortage of data scientists, specifically. Upskilling your existing non-technical teams to handle data-centric tasks is one smart way to handle the ongoing skills gap. Also, not a bad idea with threats of a recession around the corner.

4. Save Time and Money

But, a shortage of data scientists also means higher salary demands. Leveling up your current business teams to handle specialized tasks means less need to hire these ever-more expensive data scientists. Again, smart move in today’s shaky economy. Plus, don’t forget no-code/low-code UIs save all that time engineers would have spent not only writing, but debugging code.

5. Increase Satisfaction, Relieve Burnout

Think of citizen data scientists as business users with a brand-new set of shiny, high-tech tools. By adding to their skill set, the ability to build, train, and tune ML models, these SMEs enhance the impact of their specialized knowledge of business workflows. Who better to deploy ML to help boost the efficiency of your business than those who know the business inside and out? With new tools that empower them to deliver new value, these citizen data scientists also become more satisfied in their jobs.

Plus, just as Digital Workers free up your analysts to do more meaningful work, citizen data scientists free up your engineers from “burnout-inducing workload,” (which has only been exacerbated in recent years) so they can stick to more complex projects where their skills really shine.

6. Help Employees Embrace Automation

Employees understandably fear automation. One way to mitigate that is to bring them into the fold: When citizen data scientists create ML models themselves and deliver value to your org, automation becomes empowering rather than threatening. After all, your teams’ buy-in is necessary to make these program work.

Get in on the Citizen Data Scientist Action: Machine Learning Lab

This summer, WorkFusion rolled out the much-anticipated 10.2.5 release of our platform. One major new feature is Machine Learning Lab (ML Lab), a low-code/no-code environment where anyone at your org can build, train, and tune ML models from scratch. “This is something our customers have been asking for, and it’s here,” said Account Executive Ryan Buttacavoli, who was joined by VP of Product Management, Vasil Remeniuk, for a customer webinar on ML Lab and citizen data scientists.

Remeniuk provided some stats on the current state of ML projects: When building a new ML model, 80% of the time is spent creating a data set. The remaining 20% on training and tuning the model. “When we focus on that 20% alone, a lot of that is available for non-scientists—people without PhDs in computer science.”

Better Decision-Making & Faster Results

The bottom line, by bringing the AI/ML modeling closer to the business user, you can leverage their deep understanding of the tasks and jobs at hand. Essentially, a BSA officer with no AI skills can build or tune a model for decision-making in their area and can yield a stronger and faster result than having to train a data scientist on what the job and process are.

Our thinking: With the right tools, the right selection of use cases, and the right training, citizen data scientists can take on up to 80% of the work in ML model building.

To see this for yourself, watch demos of three use cases for citizen data scientists:

  1. Document Extraction [12:01–21:18]
  2. Document Classification [21:18–26:45]
  3. Spam or Not Spam Dataset [26:45–29:20]

And be sure to check out our customer webinar series, which empowers platform users with the latest product information to optimize the value of their WorkFusion program:

Episode 1: Adding Self-Service Digital Workers to Your Automation Team: A Deep Dive into WorkFusion’s 10.2.5 Release

Episode 2: So Easy an Analyst Can Do It: WorkFusion’s Citizen Data Scientist

Episode 3: Design Your Data Center: WorkFusion’s Managed Services Offering

Share this article