Coaching for HubSpot practitioners

4 sessions · 1:1 with Ryan

You're good at HubSpot. Now build on AI.

Four working sessions. By session four, you're running Claude Code, orchestrating agents, and delivering from transcript to HubSpot implementation in a single session.

Not a course. Not a demo. Live 1:1 sessions where you build the stack on your own machine — the same stack USS runs on every day.

Four 90-minute sessions · all materials · two weeks of async access after delivery · see what's inside

Live working sessions, not pre-recorded lectures You leave with working software, not notes Same stack USS uses to run the business

You configure HubSpot faster than anyone on your team. The ceiling isn't your HubSpot knowledge.

The admins who are going to matter in the next two years aren't the ones who know HubSpot best. They're the ones who know how to operate Claude Code, orchestrate agents, and turn a call transcript into delivered work without leaving their terminal.

The manual work is the ceiling.

Custom objects. Workflows. Integrations. You do the thinking; the clicking eats the day. That gap between what you can imagine and what you have time to ship is exactly where AI leverage lives.

Peers are shipping 10x faster.

Not because they work harder. Because their laptop is the operating system — Cursor, Claude Code, a working agent/skill stack, and HubSpot wired into all of it. What used to take a week takes an afternoon.

Most AI courses are demos.

You watch. You take notes. Nothing gets installed. A week later you remember the vibe but not the setup. What you actually need is someone who runs this stack daily sitting next to you while you install it.

Outcomes

What you can do after four sessions.

Run Claude Code from your own terminal — your laptop becomes the operating system.

Know when to use Claude Code vs Claude Desktop and why.

Build and modify your own agents and skills.

Query your HubSpot portal through the API and through Daeda MCP.

Go from a client call transcript to a delivered implementation in one Claude session.

Maintain a memory system that compounds across every engagement you run.

The engagement

Four working sessions. Each one installs something real.

Every session is a live working session. You bring your laptop and your portal. You leave with something running — not notes.

Foundation — The Dev Environment

Installed and working by end of session: Cursor, Claude Code, GitHub, your practitioner workspace.

  • Cursor configured and open to your first repo
  • Claude Code installed in your terminal, authenticated, ready
  • GitHub account + first private repo for your workspace
  • Folder structure that scales — clients, skills, agents, scripts
  • First commit pushed

What lands

Why the terminal matters. Why Git matters. Why your laptop is the operating system, not a tool.

Claude Architecture — Agents, Skills, MCPs

Installed and working by end of session: your first custom agent, your first skill, your first MCP.

  • First custom agent defined and running
  • First skill composed and tested
  • Claude Code vs Claude Desktop — when to use which, and why
  • First MCP server connected to your environment

What lands

The difference between prompting and orchestrating. How agents hold context. Why skills compose. How MCP changes what Claude can do.

HubSpot Integration

Installed and working by end of session: HubSpot authenticated from the terminal, live data flowing.

  • HubSpot private app created with the right scopes
  • API authenticated from your terminal
  • First live read against your portal — companies, contacts, deals
  • Daeda MCP connected for local portal querying
  • First HubSpot Project scaffolded

What lands

HubSpot from the terminal vs HubSpot from the UI. When to use the API vs the MCP. How Projects change the delivery model.

Scale — Transcript to Solution, Memory, Patterns

Installed and working by end of session: full transcript-to-implementation pipeline, memory system, delivery habits.

  • Memory system scaffolded — CLAUDE.md + memory/ structure
  • Live working example: real call transcript → scoped implementation → delivered, all in one session
  • Practitioner tips/tricks library — the habits that separate fast delivery from slow
  • Your next 90 days mapped out

What lands

How human knowledge layers into AI context. Why memory management is the skill that compounds. How to run a delivery from start to finish without leaving the terminal.

Deliverables

What you walk away with.

The engagement ends. The stack you built stays. Everything below is yours — configured, running, documented.

Fully configured dev environment on your own machine

Private GitHub repo with your own practitioner workspace

Starter agent templates you can adapt and extend

Starter skill templates you can adapt and extend

HubSpot API connection documented and reusable

Daeda MCP configured and running

Memory scaffold with CLAUDE.md and memory/ directory

Recordings of all four sessions

Private async channel with Ryan for two weeks after session four

Who this is for — and who it isn't.

HubSpot admins who have hit the ceiling of what manual work can do

Practitioners delivering client work who want to ship in hours what used to take weeks

Ops leads evaluating what their team should be doing

Anyone who watched a Claude Code session and thought "I need that running on my machine"

People who have never administered HubSpot

People looking for a passive course — you build alongside Ryan, that is the point

What you need before session one

Prerequisites

  • HubSpot admin access to a real portal (yours or a client's, with permission)
  • Mac or Linux machine preferred (Windows works with more setup)
  • Claude subscription (Pro or Team recommended)
  • Willingness to use the terminal — no prior experience required
  • 30 minutes of pre-work before session one (checklist sent 48 hours ahead)

After you book, Ryan sends a short checklist 48 hours before session one so you arrive ready.

The next step, if you want it

Mechanics get you far. Methodology gets you further.

This engagement teaches you how to operate the stack. Graduates who want to run it against a methodology — not just tools — have a natural next step through the Value-First Collective.

Mentioned here so you know it exists. Not part of this engagement, not a required path.

Ready to build on AI?

Buy now and we'll schedule session one. Or book a 20-minute call first if you want to talk it through.

$2,495 · four sessions · all materials · yours to keep.