UNLOCK YOUR DATA POTENTIAL

Data Makers Fest
Hands-on Tutorials

4^th of May 2026

Porto Business School

Embark on a learning journey tailored to your expertise level with our exclusive half-day hands-on tutorials.

These practical workshops will be led by industry experts and promise a unique learning experience that’ll make a difference in your curriculum and daily life.

Each tutorial will have a maximum of 50 participants and a duration of 3h30 (with a 30 minutes break).

The hands-on tutorials tickets are only available until the 24^th of April.

Tutorials

dbt fundamentals: an introduction to dbt in 2026

Sam Debruyn

Freelance Data Platform Architect

Sam Debruyn

Freelance Data Platform Architect

New to dbt? This hands-on tutorial is your launchpad. Starting from scratch, you'll set up your first dbt project using dbt Cloud and Snowflake trial accounts, no prior setup required. By the end of the session, you'll understand what dbt is, why it has become the backbone of modern analytics engineering, and how it fits into the broader data transformation landscape.
You'll work through the core building blocks of dbt: models, sources, seeds, and Jinja templating, before moving into macros, documentation, and testing. Every concept is reinforced with practical exercises, so you leave with working code. Whether you're a data analyst or an engineer stepping into the analytics space, this tutorial gives you the foundation to hit the ground running on your first real dbt project.

What you'll learn

- What is dbt and all of its variants
- Core dbt concepts and architecture
- Building models and defining sources
- Using Jinja for dynamic SQL
- Documenting projects and adding data tests
This should get you well on your way to start working on your first dbt project.

Expected level

Beginner / Introductory

Who is this tutorial for?

Beginning data practitioners new to dbt and analytics engineering.

Requirements for attendees

- Laptop
- Email address on which you have NOT already created a Snowflake Trial account
- SQL knowledge

Schedule

09h30 — 13h00

dbt deep-dive: beyond your first models

Sam Debruyn

Freelance Data Platform Architect

Sam Debruyn

Freelance Data Platform Architect

Already comfortable with dbt basics? This tutorial is where your skills get serious. Designed for practicing analytics engineers and data engineers who have moved past the fundamentals, this session explores the more powerful and often underused features of dbt that separate good pipelines from great ones.
Over the course of the tutorial, you'll tackle incremental models and snapshots for handling large and slowly changing data, write custom and unit tests for bulletproof pipelines, and explore Python models for when SQL isn't enough. You'll also dig into debugging strategies, discover the most valuable packages in the dbt ecosystem, and even learn how to build and publish your own. This hands-on tutorial pairs perfectly with the morning fundamentals session, but any practitioner with solid dbt experience will find it immediately applicable to their day-to-day work.

What you'll learn

- Using snapshots and incremental models effectively
- Writing custom and unit tests
- Debugging dbt runs and failures
- Working with Python models and dbt packages
You’ll gain confidence working on production-ready dbt projects.

Expected level

Intermediate / Practitioner

Who is this tutorial for?

Analytics engineers looking to level up their dbt skills. Prior dbt experience recommended.

Those that participate in the ""dbt fundamentals"" hands-on tutorial can also follow this second part.

Requirements for attendees

- A laptop
- Prior experience with dbt fundamentals (models, Jinja, macros, documentation, and testing)
- SQL knowledge

Schedule

14h30-18h00

From RAG to Agents: Implementing Agentic Search

Alexey Grigorev

Founder of DataTalks.Club and Lead Data Scientist at OLX

Alexey Grigorev

Founder of DataTalks.Club and Lead Data Scientist at OLX

Retrieval Augmented Generation put LLMs in contact with your data. Agents put them in the driver's seat. In this hands-on tutorial, you'll start by building a classic RAG system and then progressively evolve it into a fully agentic search workflow, one that doesn't just answer questions but actively reasons, explores, and decides how to find the right answer.
Step by step, you'll implement custom tools, wire them into an orchestrated agent, and learn how to evaluate and control agent behavior so your application stays reliable and useful in practice. This tutorial is intentionally accessible: if you know Python and have an OpenAI-compatible API key, you're ready to go. You'll leave with a working agentic system you built yourself and a clear mental model for where LLM applications are heading.

What you'll learn

- creating a simple RAG application
- turning RAG into a search agent
- creating tools for the agent
- testing the agent

This tutorial will give you a practical introduction to agentic search and help you understand how to structure LLM applications that go beyond simple prompt–response patterns.

Expected level

Beginner / Introductory

Who is this tutorial for?

Practitioners and developers new to RAG and agent-based LLM workflows.

Requirements for attendees

- Python knowledge
- A laptop with Python and VS Code installed
- An OpenAI API key or a compatible alternative (e.g. Azure OpenAI, Groq, xAI)

Schedule

09h30 - 13h00

How to implement SCD Type 2 with Spark and Delta Lake

André Godinho

Tech Lead & Machine Learning Engineer at DareData

José Moreno

Senior Solutions Architect at Adidas

André Godinho

Tech Lead & Machine Learning Engineer at DareData

José Moreno

Senior Solutions Architect at Adidas

Slowly Changing Dimension Type 2 is one of those patterns that every data practitioner eventually encounters and frequently struggles to implement correctly at scale. When you need to preserve the full history of how data changed over time, getting SCD Type 2 right is not optional, especially in ML pipelines where training datasets must reflect precise point-in-time reality.

This Tutorial takes a thorough, practical approach: you'll understand why SCD Type 2 matters, where naive implementations break down, and how to build an efficient, production-grade solution using Spark as the distributed compute engine and Delta Lake for ACID-compliant storage with full versioning. The session culminates in a complete implementation on Databricks, giving you code and patterns you can bring directly back to your own pipelines. Open to all experience levels, whether this is your first encounter with the pattern or you're looking to finally do it properly.

What you'll learn

In this hands-on tutorial, participants will learn:
- How to design and implement SCD Type 2 transformations
- How to use Spark for scalable, distributed data processing
- How Delta Lake enables reliable SCD implementations through ACID transactions and versioning
- How to implement and operate SCD Type 2 pipelines in Databricks

By the end of this Tutorial, you will understand how to implement SCD Type 2 in a robust and efficient way for both analytics and machine learning workloads.

Expected level

All levels / General interest

Who is this tutorial for?

Data and ML engineers working with dimensional data at scale.

Requirements for attendees

- Laptop
- Python, SQL, pandas or spark knowlege
- uv installed, Python ≥ 3.11, databricks CLI installed

Schedule

09h30 - 13h00

Partner Tutorial: Production LLM Inference on Kubernetes with vLLM

Miguel Silva

Solutions Architect at AWS

Davi Garcia

Specialist Solutions Architect - Modernization at AWS

Miguel Silva

Solutions Architect at AWS

Davi Garcia

Specialist Solutions Architect - Modernization at AWS

Deploying a large language model in a notebook is one thing; running it reliably at scale in production is another challenge entirely. This deep-dive workshop tackles exactly that, walking ML and platform engineers through the full stack of deploying vLLM-based inference services on Kubernetes in a way that's performant, observable, and production-ready.

You'll learn how to optimize GPU use, reduce inference latency by 3 to 5x, and configure horizontal scaling to handle concurrent request spikes without breaking a sweat. The session also covers distributed inference patterns for larger models, container optimization to cut model loading times, and setting up comprehensive monitoring using open-source tooling. If you're responsible for making LLMs work in the real world, not just in demos, this tutorial gives you a concrete playbook.

What you'll learn

- Deploying vLLM-based LLM inference services on Kubernetes clusters.
- Optimizing GPU utilization and reduce inference latency by 3-5x.
- Implementing horizontal scaling for handling multiple concurrent requests.
- Setting up comprehensive monitoring for LLM workloads using open-source tools.
- Applying container optimization techniques to reduce model loading times.

Expected level

Advanced / Deep Dive

Who is this tutorial for?

ML and platform engineers running LLM inference in production.

Requirements for attendees

- Laptop
- Knowledge of python, k8s, LLM fundamentals

Schedule

14h30 - 18h00

Secure your
spot now

Get your ticket!

ticket details

What’s included
in the ticket?

Hands-on learning

Engage in interactive tutorials tailored to different skill levels and interests.

Expert guidance

Learn from prominent data science and AI figures.

Comprehensive
materials

Access all necessary materials for the workshop.

Certification

Receive a certificate upon completion of the tutorial.

Refreshments

Enjoy a coffee break during the session.

Networking Lunch

Connect with peers and experts over a delicious lunch.

ticket price

How much does it cost?

Standard Workshop

1 Tutorial of your choice

375€

VAT included

Partner Workshop

AWS-hosted Tutorial

270€

VAT included

Workshop Pack

2 Tutorials of your choice

* Free conference ticket for the first 50 buyers.

Includes a free conference ticket

Enjoy the best
of the festival

Save your spot

Data Makers FestHands-on Tutorials

Tutorials

dbt fundamentals: an introduction to dbt in 2026

Sam Debruyn

dbt deep-dive: beyond your first models

Sam Debruyn

From RAG to Agents: Implementing Agentic Search

Alexey Grigorev

How to implement SCD Type 2 with Spark and Delta Lake

André Godinho

José Moreno

Partner Tutorial: Production LLM Inference on Kubernetes with vLLM

Miguel Silva

Davi Garcia

Secure yourspot now

What’s includedin the ticket?

Hands-on learning

Expert guidance

Comprehensivematerials

Certification

Refreshments

Networking Lunch

How much does it cost?

Standard Workshop

375€

Partner Workshop

270€

Workshop Pack

Includes a free conference ticket

Enjoy the bestof the festival

Partner with us

Partner with us

Brand Visibility

Employer Branding

Networking

Recruitment

Media Corner

We would love to have you there!

Become a volunteer

Why becoming a volunteer?

Meet new people

Improve your skills

Enjoy the festivall

Experience the backstage

Data Makers Fest
Hands-on Tutorials

Secure your
spot now

What’s included
in the ticket?

Comprehensive
materials

Enjoy the best
of the festival