Are You Building Pi-Shaped Teams in this AI Era?
This post was born from a frustration: the persistent myth that data engineers are merely "pipeline builders."
Let me tell you something that's been bugging me for years... This post was born out of frustration: the persistent myth that data engineers are merely "pipeline builders" – – and not true software engineers. Leaders, data scientists, and other teams also hold this generalized view. They tend to think they are limited to particular skillsets even-though they have more than that. I think this holds true for other specialized roles, like Platform Engineers as well.
Why this myth is harmful?
It feels incredibly shallow to define what we do on a day-to-day basis. Many data engineers posses expertise in machine learning, cloud architecture, product analytics and experience of building data products end-to-end.
This myth causes talent undervaluation. Often, these valuable skills go unappreciated and underutilized due to restricted job roles and narrow responsibilities. Companies are missing out.
It can stop companies from adapting to AI. If data engineers are just pipeline builders, their AI skills are ignored. Instead of using the talent they already pay for, they have to spend extra to hire more people.
Ultimately, this limits the data engineer's growth. By boxing people in and underutilizing them.
That frustration, with more thoughts given, has led me to a deeper question: How can organizations unlock the full potential of their data teams to meet the needs of a fast-evolving market, especially with all the recent AI transformations? What does the ideal role definition – look like in this new AI era?
Digital Transformation: Beyond the Silos
Let's break down digital transformation into its core pillars. I’m being very deliberate in calling this digital transformation, not just technology transformation:
Product: It's not just about the code; it’s about defining and delivering the right solutions that solve real user problems and align with business goals. It's the "what" we're building.
Insights: It goes beyond just "collecting data". it’s about understanding why users behave the way they do, what trends are shaping the market, and how we can use that knowledge to make better decisions. It's the "why" and "how".
Infrastructure: It's not just about servers and networks and cloud ; it's about building a resilient, scalable, and secure foundation that can support our entire digital ecosystem. It's the "where" and "how it scales".
Technology: It's more than just the latest frameworks; it's about choosing the right tools and languages to build and maintain our solutions effectively. It's "how do we want to get it done"
Why Silos Fall short?
As the image shows, this 4-block model ( on the left ) highlights these key areas. But, in isolation, these blocks fall short.
The reality is, value isn't created in silos. It emerges where these disciplines intersect and collaborate. A great product idea is useless without the right infrastructure to support it. The most sophisticated technology is pointless if it doesn't solve a real user need.
The world of digital transformation is constantly evolving. What was once a specialized skill becomes a baseline expectation, and today's cutting-edge expertise morphs into tomorrow's common knowledge.
To thrive in this dynamic landscape, we need to rethink how we approach skill development and embrace a model that prioritizes overlapping competencies.
The Pi-Shaped Engineer : The Rise of the Overlapping Skillset
This visually represents the core idea: that true innovation and real-world impact come from the overlap of these areas. This overlapping skillset leads us to the concept of the Pi-shaped engineer.
Now, you might already have some of these people on your team without even realizing it! While the term itself might be new to some, most people are already familiar with the idea of "T-shaped" skills but not yet the concept of being "Pi-shaped".
A Pi-shaped engineer is not just a master of one domain. They possess deep expertise in at least two distinct areas and a broad understanding of the others. This creates the "Pi" shape: two strong vertical "legs" of deep specialization with a connecting horizontal bar of interdisciplinary knowledge.
AI as the Norm in Pi-Shaped Model
Examples : The AI-Ready Pi-Shaped Roles
what was once a specialized skill eventually becomes a normal expectation, and today's "new" skills become tomorrow's expertise.
The Data-Driven Product Manager: Ten years ago, making charts and writing SQL were skills for data analysts. Now, a product manager needs those to understand data and make decisions. (Product + Insights)
The Infrastructure-Aware Developer: Just writing code isn't enough. A developer needs to know cloud basics, security, and how to make things run well using AI. (Technology + Infrastructure)
The AI-Empowered Professional: AI is specialized now. Soon, everyone will need to know it. Think of marketers using AI to personalize ads. The ability to understand and apply AI will become normal across all functions.
Why is this important
Building a Pi-Shaped Culture: A Note for Leaders
Pi-shaped engineers bring immense value, but many organizations struggle to cultivate this culture. What's getting in the way? Here are three key challenges I've identified:
The Engineer's Mindset
Some engineers may resist branching out, preferring to stay within their area of established expertise. This is often driven by a desire for deep mastery and a fear of spreading themselves too thin.
Not every company need will fit into an individuals's core interests. This misalignment can lead to disengagement or a lack of motivation to develop new skills.
Leadership Blind Spots:
Narrow Perspectives: Leaders with different experience background may not fully grasp the value of engineers who possess a broad understanding of the business, technologies, and data.
Lack of Support and Recognition: Without active leadership support, engineers may struggle to pursue cross-functional development, lacking the resources, time, or opportunities to expand their skillsets.
Glorifying Visibility Over Substance:
All that glitters is not gold : companies sometimes overvalue what they can see—such as polished UIs, flashy AI demo and presentations—while undervaluing the deep, invisible work like data architecture, infrastructure, and backend systems that are essential for long-term success.
Building Your Pi: Note for Engineers
Know Your Strengths: What are you already really good at? What comes naturally to you? (This is your first "leg".)
Find Your "Adjacent": What other areas connect to your strengths? What skills would make you even better at your core role? What interests you? (This is your second "leg".)
Never Stop Learning: Take courses, go to workshops, try side projects.
Get Involved: Volunteer for different team projects to see new things.