Practical AI Podcast: "The AI Engineer Skills Gap"

Date: December 10, 2025
Host: Daniel Whitenack (Prediction Guard)
Co-host: Chris Benson (Lockheed Martin)
Guest: Ramin Mohammadi (Northeastern University & ibaset)

Episode Overview

This episode dives deep into the changing landscape of AI and data science careers, focusing on the widening skills gap between academic training and the evolving demands of industry. With the rise of GenAI, the shift in what it means to be employable in AI, and the role of practical versus theoretical education, the hosts and guest discuss how both academia and industry must adapt. Ramin Mohammadi brings a dual perspective as both an educator and a principal AI engineer.

Key Discussion Points & Insights

1. The Evolution of the AI/ML Job Market

• Data Science’s Early Appeal:

  • The role was once “the sexiest job of the 21st century” (Harvard Business Review, 2012) [02:48].
  • Universities launched new programs rapidly to feed job market demand.

• Changing Demand & the New Hiring Bar:

  • “The promise was pretty simple. Get a degree and learn a little bit of machine learning and you also, you're instantly employable. That promise feels like almost like a myth now.” — Ramin [03:24]
  • Entry-level roles now expect mid-level skillsets: building, deploying, and maintaining real, scalable AI systems [04:20].

• The Myth of the 'Full-Stack' AI Engineer:

  • Today’s AI professionals are expected to master modeling, software engineering, infrastructure, and statistics [04:20–05:35].
  • Pressure to be "full-stack" is widespread but often unrealistic.

2. How AI & Automation Disrupted Traditional Career Ladders

• The MLOps Revolution:

  • "Successful ML needs a suite of real engineering skills... containerization with Docker, CI/CD pipeline automation, monitoring.” — Ramin [05:35]
  • AI roles turned from analysis in notebooks to engineers owning pipelines and deployments.

• Generative AI’s Impact:

  • Generative AI triggered a “tidal wave," automating many tasks that were entry points for early-career professionals.
  • “Any repeatable tasks that used to be given to juniors…are highly vulnerable to AI and innovation.” — Ramin, referencing an OpenAI/UPenn study [08:42].
  • The economic case for hiring many juniors has weakened; organizations prefer a few hires with proven capabilities [08:42–09:30].

3. The Growing Academia-Industry Divide

• Educational ‘Bottleneck’:

  • Academic programs focus on foundations — math, theory, research — but rarely bridge to practical implementation [12:45].
  • “The curriculums often just stop there...there’s this huge gap between what the student learns and what employers actually need.” — Ramin [12:45]
  • Updating university courses is slow; tools and standards evolve much faster in industry than in academia [15:30].

• Industry Is Leading AI Innovation:

  • “Right now, about 70% of AI PhDs are just skipping academia and go to the job market.” — Ramin citing MIT study [15:45]
  • “96% of the major state-of-the-art systems comes from industry labs, not from universities anymore.” — Ramin [15:45]

4. Innovative Academic Approaches to Closing the Skills Gap

• Hands-on, Product-Focused Courses:
  • Ramin developed an MLOps course at Northeastern with 150–270 students per class.
  • Students pick a domain, work in teams, and build a real product for an industry expo [19:35–22:10].
  • “You don’t just learn ML anymore, we teach you how to build with it.” — Ramin [21:09]

5. The Double-Edged Sword of GenAI for Students and New Hires

• GenAI as Leveler and Divider:
  • “The portfolio has become the new credential.” — Ramin [29:06]
  • Students self-organize learning: 60% take online courses, 82% participate in hackathons [29:06].
  • But hands-on cloud experience is expensive, giving an edge to those with more resources — “the bar isn’t just higher, it’s financially more expensive for students to learn.” [30:39]

6. Adapting as a Senior AI Engineer

• Shifts at All Levels:
  • Even senior/principal AI engineers must continually adapt: “I need to…work a lot with LLMs...have a better understanding on…GPU optimization…” — Ramin [32:26]
  • The pace of change affects both new graduates and seasoned professionals.

7. The Responsibility for Closing the Gap

• Industry-Academia Collaboration Needed:

  • “At some point, industry needs to help academia…otherwise it’s like chasing a ball; academia just constantly trying to keep up and that's not going to win.” — Ramin [37:43]

• Advice for High Schools & Early Learners:

  • Practical, project-based exposure, not just theory, from a younger age.
  • “There are lots of AI capabilities that you don't technically need the math behind them. You can just build a system just by knowing how to put the components together.” — Ramin [39:33]

Notable Quotes & Memorable Moments

• On Job Market Brutality:

  • “The market's absolutely brutal…it's not about what you know from the textbook anymore, it's about what can you build, can you deploy and maintain a real scalable AI system?” — Ramin [03:24]

• On the Evolving Bar:

  • “The new entry-level jobs is technically what we would call mid-level engineers a couple of years back.” — Ramin [10:55]

• On Self-Driven Learning:

  • “The portfolio has become the new credential. It's no longer about your grade, it's like about what you have as a portfolio.” — Ramin [29:06]

• On the Pace of Change:

  • “By the time university approves one new course…tools have already changed three times.” — Ramin [15:30]

• On Academia’s Need for Industry Support:

  • “Industry needs to help academia…otherwise it's just academia trying to catch up, which is not going to work.” — Ramin [37:43]

• Practical Advice for Early Learners:

  • “From high school, understand the concept, not maybe the master theory behind AI, but just to learn in general what, how does AI work?” — Ramin [39:33]

Timestamps for Important Segments

• 02:48 — The historical context of data science as a hot job and how perceptions changed
• 05:35 — Emergence of MLOps and the elevation of required engineering skills
• 08:42 — Impact of automation on junior roles and hiring strategies
• 12:45 — The educational bottleneck: theoretical grounding vs. practical readiness
• 15:45 — Industry outpacing academia in pushing state-of-the-art AI
• 19:35 — Effective academic innovations: Northeastern’s hands-on MLOps course
• 29:06 — Student strategies in bridging the skills gap on their own
• 32:26 — How senior professionals are adapting to the new demands of AI roles
• 37:43 — The call for greater industry involvement in academic preparation
• 39:33 — Advice for high schoolers and early learners wanting to enter the AI space
• 42:13 — Personal excitement about robotics and new AI domains for 2026

Tone & Style

The conversation blends candid realism with optimism, balancing concerns about access and fairness in AI career preparation with creative ideas for bridging gaps—delivered with humor and camaraderie among the speakers. Ramin’s dual perspective as an educator and engineer grounds the episode, while Chris and Daniel provide practical industry counterpoints and personal anecdotes.

Conclusion

This episode provides an unvarnished look at the escalating skills gap for AI engineers, shaped by rapid technological change and slow-moving curricula. While the challenges are significant, the guests spotlight creative academic solutions and call for more industry involvement to support equitable, practical learning. The future of AI work, they agree, will require both old and new skills—with a strong emphasis on hands-on experience, adaptability, and collaboration between academia and industry.