Context Engineering for Long Analytics Sessions

Hosted by Gui Diaz-Berrio

Tue, Apr 14, 2026

11:30 AM UTC (45 minutes)

Virtual (Zoom)

Free to join

Invite your network

Go deeper with a course

Marketing Science Bootcamp: MMM, Attribution and Experiments with Claude Code
Gui Diaz-Berrio
View syllabus

What you'll learn

Understand why long sessions break Claude Code

Context rot is real: model specs get confused, earlier decisions are forgotten. Learn what causes it and how to spot it

Learn context engineering techniques that work

CLAUDE.md design, Skills, session structure. See the patterns that keep Claude Code consistent across a 2-hour analysis

Apply this to your own analytical workflows

Walk away with a framework for structuring any long session, so Claude Code stays useful instead of becoming unreliable

Why this topic matters

A 10-minute Claude Code demo looks magical. A 2-hour analytical session exposes the cracks — wrong model specs, forgotten decisions, drifting outputs. Context engineering is the skill that separates impressive demos from reliable workflows. This lesson shows you how to structure sessions so Claude Code stays sharp from start to finish.

You'll learn from

Gui Diaz-Berrio

Led measurement for €100M+ in ad spend at Kindred Group & Just Eat Takeaway

What you'll see in 30 minutes is one session's worth of material.


The full 4-week cohort (starting May 2026) covers the complete stack:

  • Context window management for long analytical sessions
  • Building custom Claude Code Skills that encode your team's modeling standards
  • End-to-end data pipelines — from raw data to executive-ready deliverables
  • Marketing Mix Modeling and Geo-Experimentation, orchestrated agenically
  • Automating recurring measurement workflows so they run without you


Join the waitlist to get early access and priority enrollment for the May cohort of Marketing Science Bootcamp with Claude Code.


Gui Diaz-Berrio has led marketing measurement at companies spending €100M+ on advertising, including Kindred Group (Head of Marketing Analytics) and enterprise clients through Pinemarsh Consulting. He's built models from scratch and managed vendor relationships — both perspectives taught in this course.

Author of "Data Analytics for Marketing with Python" (Packt, 2024) — covering the practical frameworks taught in this course.

He's faced the challenges you're dealing with — messy data, skeptical CFOs, vendor black boxes, and the pressure to prove ROI with limited experimentation budget.

Previously at

BMW Group
FDJ United
Just-Eat.co.uk
@Packtpub

Sign up to join this lesson

By continuing, you agree to Maven's Terms and Privacy Policy.