 |
| Original Post in LinkedIn |
Hello all, I have been reflecting a lot while preparing the next semester.
AI skills are rising fast among students. But deep hands-on data engineering skills are decreasing. And that worries me.
So I decided to change the way I teach Big Data.
This year I will teach:
• One program in English (39 students)
• One program in Spanish (17 students)
Same approach but evolving the structure significantly. As I mentioned before there is a clear trend: Students are becoming increasingly AI-proficient… but less hands-on in core data engineering and platform fundamentals.
So I redesigned the course.
The structure now includes:
• 18 hours of theory
• 14 hours of guided hands-on lab
• 8 hours of stand-ups (individual Big Data topics + group final use case presentation)
But the real shift is deeper. The lab is no longer tool-driven. It is architecture-driven.
Students will learn by doing:
– Agile SAFe methodology
– GitHub & GitHub Organizations
– Terraform (Infrastructure as Code)
– AWS fundamentals (Not so agnostic but key to learn ;) )
– Spark & Spark Cluster (Master/Workers)
– Batch & Streaming ELT with Spark
– Data Lakehouse modeling (Medallion naming convention architecture)
– Governance conventions & naming discipline
– LLM + RAG + Agents over the data they built
All deployed in a controlled cloud-simulated environment using Docker. Vendor-agnostic. Enterprise-grade. Architecture-first.
The goal is simple, prepare professionals capable of becoming:
• Data Platform Architects
• Data Engineers
• Cloud Data DevOps Engineers
• AI Engineers grounded in data discipline
AI without data architecture is fragile.
Data engineering without governance does not scale.
Education must evolve with the industry.
This semester, we go deeper.
— Manuel
Comments