Data Science 2.0 with Suzanne El-Moursi from Brighthive |🎙️#63
Data Science 2.0 is the shift from dashboards to decision workflows. Suzanne El-Moursi shows how teams can go beyond “reports nobody trusts” to agentic, governed analytics, using a pragmatic stack (infrastructure → composable tools → governance) so insight can flow from ingestion through modeling, validation, and visualization without the usual stalls. We focus on how the data scientist role evolves from dashboard factory to orchestration and stewardship so the right decision happens at the right time. Also in this episode:
- Problems we face in data in different eras of tech;
- The role of a data scientist in the age of AI;
- What parts of the classic data scientist job fade?
- Who does the data evaluation?
- 3 things to learn right now;
- How to make your project AI agent ready?
You can listen to episode 63 of DevOps Accents on Spotify, or right now:
Over the past three decades, Suzanne El-Moursi has witnessed the evolution of technology—from the birth of the internet to the mobile and cloud eras—and with it, the persistent struggle of working with data. In a conversation with Leo, she shared how these challenges shaped her journey and why she believes agentic AI is the turning point for making data work more efficient, trustworthy, and transformative.
The Persistent Data Problem Across Eras
From the early days of digitizing insurance applications at IBM to the explosion of SaaS platforms and the modern data stack, Suzanne has consistently seen one issue: data is fragmented, siloed, and difficult to work with. Every technological leap—from web browsers to mobile apps—created new ways of collecting and consuming data, but also added new layers of complexity.
Leo noted how many organizations he encounters are still “data rich but insights poor,” a sentiment Suzanne echoed. Even as tools advanced, the workflows around data remained cumbersome, leaving teams with dashboards no one trusted and reports that often collected dust.
This is not a new problem. The reason it is that difficult is because throughout time technology is evolving, but the working with data problem is not improving. And more and more data is being produced. In the last two years since the rise of AI from demand surge, I read the statistics: 90% of the world's data was produced just in the last two years. Every new technology era drives an explosion of data creation, and it's upside down. We're creating more platforms that are very capable, very powerful, but we're also adding to the debt of what it takes to work and organize and have clean data assets to fuel. — Suzanne EL-Moursi
The Shifting Role of the Data Scientist
Suzanne is quick to clarify that she doesn’t see herself as a traditional data scientist. While data scientists have long been burdened with grunt work—cleaning, moving, and validating data—she believes the profession is at a crossroads.
Leo framed it well: in agentic systems, “context beats training.” Rather than endless fine-tuning of models, the real value comes from sense-makers who can provide agents with the right organizational knowledge and guardrails. Suzanne agreed, suggesting that the data scientist role is evolving toward redefining workflows, optimizing context, and ensuring data can be mined more efficiently. This shift liberates professionals from mundane tasks, allowing them to focus on innovation and creativity.
You can't just be a data scientist. That's how it changes for you. You have to understand the process and the workflows inside these companies because these agents that we're saying are going to work alongside you, and it's no longer they may, they will. The evolution and adopting of technology will be where companies are just going to have agentic work. It would make business sense to have an agentic workforce alongside humans. So then how do they coexist? The workflow is point number two. The workflows inside companies, large or small, and in the middle, any organization, these workflows are defined by people. — Suzanne EL-Moursi
Who Owns Data Evaluation?
When asked who decides what data and context are “good enough,” Suzanne highlighted a striking reality: only 2–3% of employees in most organizations are data engineers or scientists. The other 97%—the business users—are the true data consumers. They hold valuable knowledge about processes, customers, and markets, making them essential in shaping context for AI systems.
This means the responsibility for data evaluation can’t rest solely with technical teams. Instead, organizations must upskill and retrain their broader workforce, equipping them to work alongside AI agents and contribute their domain expertise.
The business, 97%, those are the data consumers. Those are the ones that are responsible for all of this because the 3%, the 2–3%—and by the way, that's if you can find data scientists. As you know, this trade is scarce. There's more demand than supply. That's why that number is low. And so you're overburdened by a lot of requests from the business for the dashboards that nobody trusts, for the dashboards that are never updated, for the reports that are always collecting dust. The value creation is not optimized for. And that 97% that is in those organizations, who are very smart people, just not in the data science space, but who have years of experience in the industry, in the business, in the exact company they're in, in the market—those are the people we are speaking about to be equipped to help bring that context and optimize the context in the workflows in these large and medium companies. — Suzanne EL-Moursi
Three Skills to Learn Right Now
Suzanne offered three practical capabilities for anyone looking to stay ahead in this new era:
- Think in terms of efficiency – Before starting any task, ask: what’s the most efficient way to do this, given today’s tools?
- Master prompt engineering – The new “English” is learning how to communicate effectively with AI systems.
- Embrace velocity in work – Use AI tools not just to complete tasks, but to accelerate workflows while maintaining quality.
Leo added that those who thrive are often the ones who act like librarians—curating knowledge, making their work trustworthy, and narrating decisions clearly for both humans and machines.
Prompt engineering is the new English. It really is. It's the bridge. It's the blending of both. Our children—I have three children—and I just imagine them. We had to learn the user interface, the GUI. What's going to be natural to them? There will be, "What's GUI, Mom?" What's going to be for them is prompting and hearing. Because that is the mechanism, that's the glue between us and this intelligence that we're talking about. That's the interaction medium. And that still allows for human genius to exist. Because all of this prompting, the wording, the style, the grammar—that's going to evolve. I'm so excited for how that world is going to evolve. Is that going to be a world where we say alphanumeric and laugh at that? What is prompt engineering? That's the new English. — Suzanne EL-Moursi
Preparing Projects for Agentic AI
Suzanne and Leo discussed the “three-layer cake” of modern data work: a composable stack as the base, agentic AI in the middle, and governance as the top layer. While agents can now perform entire data lifecycles—from ingestion to visualization—the hardest challenge remains governance.
According to Suzanne, companies often rush to build tools and agents but underestimate the importance of controls, contracts, and trust. The true measure of readiness is not just technical capability, but whether users find the system both delightful to use and trustworthy in how it handles data.
Show Notes
About Suzanne El-Moursi: Co-founder & CEO of Brighthive, a Chicago-based platform building agentic AI for the data lifecycle (BrightAgent) with governance built in; earlier leadership stints at IBM and GE Healthcare; and a board member at Uniting Voices Chicago.
- Brighthive Newsletter on LinkedIn where they share their expertise and observation on the Data Science transformation.
- follow Suzanne on LinkedIn for more insights.
- Check out Brighthive, Suzanne's company, an AI-powered enterprise data platform that uses coordinated agents to automate the full data lifecycle inside a data warehouse.
Podcast editing: Mila Jones, milajonesproduction@gmail.com