Lessons From Udemy CEO on Turning AI Hype Into Impact

Key Takeaways:

• In a Q&A interview, Udemy CEO Hugo Sarrazin shares actionable strategies for employers on how to prepare teams for AI-driven change. 
• Sarrazin recommends a shift from chasing “perfect models” to skills-first execution by targeting priority capabilities (like AI fluency), linking them to measurable outcomes and iterating fast.
• Put AI where it delivers now: automate workflows (like summarization, standard intake) while leaders and experts still design the work upfront and own QA.

Over the past year, there's been a tremendous amount of investment in new ventures — but nearly half of those investments have gone to just 10 companies. “This demonstrates how people are shifting to fewer tech stacks,” says Hugo Sarrazin, President and CEO of AI-powered upskilling platform Udemy. “If you're not thinking about what this means two or three years from now, you'll be playing defense — which isn't a good position.”

In this edition of Indeed’s Q&A series with talent leaders shaping the future of work, Sarrazin offers pragmatic guidance on preparing teams for AI-driven change. As a former global program leader at McKinsey & Co. and a contributor to Indeed’s latest Smarter Hiring Report, he takes a grounded, data-informed approach to his leadership, coupled with a human-first philosophy. 

In a new era of AI capabilities, Sarrazin’s message is simple and urgent: This is a transformational moment, and everybody needs to lean in.

📈 For more advice from Sarrazin and other workforce leaders, download the free Indeed Smarter Hiring Insights report. And don’t miss out on the latest insights: sign up for our newsletter to get new Q&As and data delivered straight to your inbox.

Meet Hugo Sarrazin

Role: Udemy CEO and President as of March 2025

At a glance: “The real value of a skills-first approach isn’t in a ‘perfect’ skill ontology — it’s solving business problems. Pick a priority and build a practical skill journey tied to measurable outcomes. Hit those goals, then move on to the next.”

Our work with Indeed is incredibly strategic,” says Sarrazin. Through Indeed’s partnership with Udemy, job seekers new to the upskilling platform can access courses in technical, business and soft skills, and then showcase those specific abilities on their resumes.

Q1: Coming into your new role leading a company that needs to adapt for an uncertain future, what’s top of mind for you?

Sarrazin: Fifteen years ago, the online learning space was a significant innovation in access to content. But today’s technology is allowing us to achieve learning outcomes in ways that weren’t economically feasible before. My top priority is to help clarify what this new category can become, where it's going and what capabilities we need. 

For example, today’s online learning and streaming video models can only deliver one-size-fits-all content — regardless of prior knowledge, neurodiversity or learning styles. But with AI, we can offer personalized, multimodal learning, which research shows can lead to an improvement in learning outcomes. AI can assess what you and I know at the start, how you learn versus how I learn, then assemble and adapt the experience from long-form to short-form content, labs, role play and so on.

Beyond AI, this creates opportunities to be part of a rich ecosystem and collaborate fluidly with partners, like Indeed. By exchanging the right information while respecting privacy laws, we can create many opportunities for people to find the jobs they deserve and help them grow.

🏗️ Talent Strategy Play: Shift from one-size-fits-all to AI-personalized learning that delivers measurable performance gains.

Q2: Indeed’s Smarter Hiring Report finds 40% of employers identify leadership and workforce skills as critical skills gaps. Where should leaders prioritize their L&D investments: AI or leadership skill building?

This is really a false choice. The most successful organizations recognize that AI fluency and leadership development aren't competing priorities; they're complementary imperatives. 

According to our Global Learning and Skills Trends Report, AI-related course enrollments have grown fivefold year over year globally, yet we're simultaneously seeing unprecedented demand for adaptive skills like critical thinking, strategic decision-making and emotional intelligence. The reality is that AI is making distinctly human leadership capabilities more premium, not less. 

AI is making distinctly human leadership capabilities more premium, not less.

— Udemy CEO Hugo Sarrazin

Smart L&D leaders are investing in both. They’re building AI fluency so their teams can leverage these powerful tools effectively. Simultaneously, they’re doubling down on the leadership and communication skills that become even more critical when humans need to guide AI-augmented workflows, make complex decisions and inspire teams through rapid change. 

At Udemy, we're seeing the most successful organizations create integrated learning pathways that develop both technical AI capabilities and the adaptive leadership skills needed to thrive in an AI-powered workplace — because the future belongs to leaders who can seamlessly blend human judgment with AI acceleration.

💡 Indeed Tip: For more on leadership, check out Indeed Chief Marketing Officer James Whitemore on Udemy’s podcast, “Leading Up: The Work Shift.”

Q3: What kind of hiring and upskilling challenges do you have among your own workforce?

We've made the pivot to a skill-based organization, so job descriptions are now less about degrees and more about specific skills. We have a dedicated day where we encourage everyone to spend time on a learning journey of their choice. Some activities are application-based, while others focus on pure learning.

Currently, we're adapting to the rapid pace of change brought on by AI. We're running internal programs to train our leaders on how to empower and work as cross-functional leaders. These are six-week programs with 20 to 30 people in each cohort. Participants go through a systematic program and demonstrate what it means to operate with mindsets, norms and behaviors that will be important for the company moving forward.

🏗️ Talent Strategy Play: Codify skills-first now — map roles to skills, assess for them in your workforce and protect time for learning.

Q4: We find that a lot of leaders and companies are interested in skills-first hiring but don't quite know how to get started. Why was it important to make the pivot to skills-first hiring, and what were the most important steps in making that transition?

At a macro level, the reason is similar to what everyone else is experiencing: We need to adapt our organizational model to the realities of the 21st century, where different expertise and skills are essential for various jobs and roles. And we need to be more deliberate in codifying and identifying these skills. 

This approach allowed us to expand our talent pool by being more creative and intentional about which skills we seek. Internally, it allowed us to be much more explicit in writing performance reviews and defining expectations. Last, it aids in clarifying different career paths, which helps us retain employees.

We have a dedicated day where we encourage everyone to spend time on a learning journey of their choice.

— Udemy CEO Hugo Sarrazin

One area I’m passionate about is how organizations see the promise of skills-first, then rush to codify everything, sparking debates over skill ontologies. Tool A uses one, Tool B another, and suddenly we’re mapping frameworks instead of making progress. While interesting, it defeats the purpose.

The real value isn’t in a “perfect” skill ontology — it’s solving business problems. Pick a priority — say, company-wide AI fluency — and build a practical skill model tied to measurable outcomes. Hit those goals, then move on to the next.

🏗️ Talent Strategy Play: Skip ontology debates; start with one business-critical skill (ex. AI fluency) tied to KPIs, then scale.

Q5: As a leader of a large company, what's one big idea you think is unfamiliar but necessary to adapt to the changing future of work?

The most important capability is the ability to change. This level of innovation will create an incredible amount of disruption across industries in terms of how work will be done.

To get more specific, AI is a probabilistic model trained on a vast data set, but it has no true intelligence, so it can produce some odd results. For example, a data scientist found that if you ask different AI models to generate a random number, most are likely to say 27 — indicating a learned bias.

Now, if we look at jobs that are unstructured and historically hard to automate, AI allows us to start automating them — tasks like summarization, transformation and generation. I define this work in three phases: design, execution and quality control. The execution is less time-consuming with AI, but quality control still requires human judgment, insight and nuance. What we need now are people who can think creatively about designing the work upfront, outsource work and maintain oversight and end-stage review.

🏗️ Talent Strategy Play: Model work after design–execution–quality control stages. Use AI to assist with execution, but keep humans accountable for quality.

Indeed provides this information as a courtesy to users of this site. Please note that we are not your recruiting or legal advisor, we are not responsible for the content of your job descriptions, and none of the information provided herein guarantees performance.
_____

Related Links 

Inside Salesforce’s AI-First Leadership With Agentforce COO Madhav Thattai

Adam Grant’s Warning to Leaders: Your Culture Carriers Are Running on Empty