The AI-First Data Engineer: 10–50x Productivity and What Changes Next
Get the full intelligence
Search transcripts, export clips, track mentions, and explore all topics from “The AI-First Data Engineer: 10–50x Productivity and What Changes Next” inside PodZeus.
In this episode of the Data Engineering Podcast, host Tobias Macy welcomes back Gleb Majanski, CEO and co-founder of Datafold, to discuss the transformative impact of AI on data engineering in 2026. Gleb shares his firsthand experience transitioning from a traditional data engineer to an 'AI-first' practitioner, highlighting how agentic coding—where AI agents autonomously write, execute, test, and debug code—can boost productivity by 10 to 50 times. He contrasts this with basic AI-assisted coding, emphasizing that true agentic workflows represent a paradigm shift, turning data engineers into 'drivers' of autonomous systems rather than manual coders. The conversation explores the implications of this shift: the decline of commodity skills like manual SQL writing, the rise of cross-functional 'product-minded' data professionals, and the consolidation of the fragmented data stack into AI-native platforms. Gleb also addresses critical challenges such as data privacy, the need for secure agent deployment within enterprise security perimeters, and the importance of context—especially through tools like data knowledge graphs—to enable AI to reason effectively across complex data ecosystems. He concludes with practical advice: modernize legacy infrastructure, experiment with AI tools, build internal AI workflows, and embrace the fact that mastering AI is now a core craft, not a shortcut.
Agentic coding—where AI agents autonomously write, execute, test, and debug code—can increase data engineering productivity by 10–50x.
The role of the data engineer is shifting from code writer to operator of AI agents, requiring stronger business acumen and product thinking.
Legacy data infrastructure is a major barrier to AI adoption; modern data platforms (Databricks, Snowflake) are essential for secure, AI-native workflows.
Data quality is no longer about human-curated tests; it's about enabling AI to reason across all data, including imperfect data, with proper context.
Data teams should build their own AI workflows and tools, as the best solutions are emerging from experimentation, not pre-built platforms.
The Bottleneck of Data Engineering
Tobias introduces the pain point of data teams being overwhelmed by constant requests for dashboards and reports, setting the stage for AI-driven solutions.
Introducing Gleb Majanski and the AI Awakening
“I was completely blown away. I always thought of myself as someone who can write really good SQL and in my day at Data Engineers, I think that was a superpower and that's how you advanced in your career and that's how you got things done. With the current capabilities of agentic coding, I just thought that this completely changes the job and the experience.”
Defining Agentic Coding vs. Basic AI Assistance
“An agent would actually be able to not only write the code for you but execute that code against the database, get the results, evaluate the results, put the code into, let's say, dbt model, run dbt, debug it, write tests, debug it, and then present you with the complete outcome.”
The Security and Privacy Challenge of AI in Data
Gleb addresses enterprise concerns about AI accessing sensitive data, explaining how using AI within cloud data platforms (like Databricks or Snowflake) keeps data and models within the same security perimeter.
The Future of Data Engineering: From Code to Outcomes
“The demand for data engineering as a way to deliver high quality data to power data driven decisions is going to actually grow. And in economics, there is this famous Jevons paradox that essentially says that if the price for a given resource capability drops, we'll actually see more of that being consumed.”
“I was completely blown away. I always thought of myself as someone who can write really good SQL and in my day at Data Engineers, I think that was a superpower and that's how you advanced in your career and that's how you got things done. With the current capabilities of agentic coding, I just thought that this completely changes the job and the experience.”
“What matters in the AI world is that A, AI has ability to act. So we talked about agentic loop, execute queries, evaluate queries, run tools, run tools like dbt and two, AI has access to all of your data because again, the more data points you have, the more complete picture you can construct.”
“An agent would actually be able to not only write the code for you but execute that code against the database, get the results, evaluate the results, put the code into, let's say, dbt model, run dbt, debug it, write tests, debug it, and then present you with the complete outcome.”
Host
Guest
Datafold
organization
Gleb Majanski
person
Tobias Macy
person
Databricks
organization
dbt
product
Snowflake
organization
Cloud Code
product
Retool
organization
Superset
product
GitHub Copilot
product
Get the full intelligence
Search transcripts, export clips, track mentions, and explore all topics from “The AI-First Data Engineer: 10–50x Productivity and What Changes Next” inside PodZeus.
Start discovering podcast insights today
Start with a 7-day trial and explore a growing catalog of popular podcasts. No credit card required.
No credit card required • 7-day trial • Cancel anytime