Skip to content
Shows/Human + Machine
Human + Machine·Episode 5·Series

Giving AI the Context It Needs

Context engineering as the real skill

May 21, 2026·34:00·Daniel Rodriguez, Paul Maxwell

Video coming soon

No YouTube video has been added yet.

Your browser does not support the audio element.

About this episode

The gap between mediocre and remarkable AI output is almost always context. Paul and Daniel dig into context engineering: what to include, what to leave out, and how to build reusable context that pays off every session.

Key ideas

  • 01Context, not clever wording, drives output quality
  • 02More context isn't better; choose what matters for the task
  • 03Reusable context packs pay off every session
  • 04Too thin and the model guesses; too noisy and it gets distracted

Show notes

What we covered

If prompting is the visible skill, context engineering is the one that actually moves quality. Paul and Daniel go deep on how to assemble the right context for a task, why more isn't always better, and how to build context assets you reuse instead of rebuilding every time.

In this episode

  • Why context, not clever wording, drives output quality
  • What to include and what to deliberately leave out
  • Building reusable context packs for recurring work
  • The signs your context is too thin or too noisy
  • Treating context as an asset you maintain

Links & resources

  • Claude + HubSpot Context Pack (downloadable)
  • Context assembly checklist
  • Daniel on maintaining context as an asset

Timestamps

Chapters

  1. —Context over wording
  2. 5:40What to include and exclude
  3. 13:40Building reusable packs
  4. 22:20Too thin vs. too noisy
  5. 28:40Context as a maintained asset

Quotable

“Think of context as an asset. Build it once, and it pays off every session after.”

— Paul Maxwell

Full transcript

Transcript~1 min read

Show transcriptHide transcript

Paul: The difference between mediocre and remarkable output is almost never the wording of the prompt. It's the context. What does the model actually know about your situation before it starts?

Daniel: And more isn't automatically better. Dump everything and you drown the signal. The skill is choosing what matters for this task and leaving the rest out.

Paul: Think of context as an asset. You build a good pack once, the schema, the conventions, the examples, and it pays off every single session after that.

Daniel: Too thin and the model guesses. Too noisy and it gets distracted. Context engineering is finding that middle, and it's a real, learnable skill.

Mentioned

  • Claude
  • HubSpot
  • Anthropic

Related links

  • Context engineering fundamentals

Hosts

  • DR
    Daniel Rodriguez

    Host

    RevOps Director, RevOps HQ

  • PM
    Paul Maxwell

    Host

    CEO, RevOps HQ

Topics

  • AI Tooling
  • Artificial Intelligence
  • Context Engineering

Downloads & resources

  • Claude + HubSpot Context PackTemplate

Subscribe to Human + Machine

  • YouTube
  • Spotify
  • Apple Podcasts

Related episodes

Human + MachineAI Slop Is Your FaultPaul and Daniel make an uncomfortable case: most AI slop comes from lazy prompting and missing context, not the model. They show how to take responsibility for quality and get dramatically better results.
Human + MachineHow We Use Claude in Our WorkNo theory this time. Paul and Daniel open up their actual day-to-day: where Claude saves them hours, where they deliberately keep it out, and the habits that keep the work high quality.
Human + MachineThe Limits of Claude's HubSpot ConnectorConnectors make AI-to-CRM feel seamless, but every integration has edges. Paul and Daniel map what Claude's HubSpot connector handles gracefully and where you still need custom tooling.
Human + MachineUsing Claude with HubSpotPaul and Daniel walk through concrete ways they use Claude alongside HubSpot, from drafting to analysis to cleanup, and the context pack that makes the results reliable instead of hit or miss.