Skip to content
Lance Wills
/ sheet 03 — the inventor's lab

AI + Code: Building the GTM Systems I Used to Wish Existed.

Agentic workflows, attribution infrastructure, CRM automation, content systems, and experiments built with modern AI tooling.

note to readerI am not positioning myself as a pure software engineer. I am a GTM leader who builds enough to prototype systems, shape product direction, automate workflows, and work credibly with technical teams.

/ lab 01 — featured build

RAIL — Growth Intelligence + Attribution System

RAIL connects marketing activity, product usage, CRM stages, meetings, pipeline, and revenue into a clearer operating layer for GTM teams.

working prototype — not a slide
  1. Ad spend
  2. UTM
  3. Website
  4. Product usage
  5. CRM stage
  6. Sales opp
  7. Revenue
  8. Reporting
fig. — one continuous rail from spend to revenue. no orphaned data.

Feature surface

  • UTM Builder
  • Attribution chart
  • Forecasting
  • Marketing Pipeline
  • Buying Committee
  • CRM workflow integration
  • Gmail / Calendar meeting signal capture
  • Ad platform imports
  • AI recommendations (future-state)

Stack

  • React
  • Tailwind
  • Convex
  • Twenty CRM
  • Google Ads
  • LinkedIn Ads import
  • Reddit Ads (planned)
  • Gmail / Calendar signal capture
  • HubSpot / CRM concepts
  • AI-assisted workflows
/ lab 02 — agentic GTM workflows

Agents on the Payroll.

Each workflow removes a manual GTM task, captures a signal a human would miss, or raises the quality floor of the output.

/ agent 01

Signal Capture Agents

Capture meaningful buyer signals from email, calendar, LinkedIn, web activity, campaign engagement, and CRM movement.
/ agent 02

Content Quality Pipeline

Turn seed topics into stronger LinkedIn/X posts using ICP-specific scoring, motivation dimensions, technical/operator proof, and iterative refinement.
/ agent 03

CRM Workflow Automation

Create pipeline opportunities from meetings, update stages, enrich context, and reduce manual GTM ops work.
/ agent 04

Research + Briefing Agents

Generate account, category, ICP, and messaging research for technical products and GTM planning.
/ agent 05

Attribution + Reporting Agents

Connect channel activity to pipeline and revenue, then surface what changed, why, and where to act next.
/ lab 03 — bench experiments

The Experiment Log.

~/growth-lab — claude code
❯ claude
// wiring gmail + calendar signals → twenty crm
✓ meeting detected: acme corp — technical eval
✓ opportunity created: stage = discovery
✓ attribution: linkedin ads → utm_campaign=q3-rag
// next: score content draft against icp gates
quality score: 7.9 / 10 — revise hook, keep proof
❯ _

Representative session — the lab notebook, in terminal form.

Claude Code builds

Agentic coding as a GTM prototyping tool

Local LLM experiments

Qwen · Ollama · Aider comparisons

Prompt systems

Reusable, ICP-aware prompt architecture

GitHub repos

Version-controlled GTM and automation builds

Marketing automation prototypes

Workflow + trigger design

LinkedIn/X workflow

Idea → quality gates → publish loop

Dripify → CRM signal capture

Outbound engagement into pipeline

Twenty CRM architecture

Open-source CRM as a GTM data spine

GitHub

Selected AI, GTM, and automation builds.

View Repos

The Thinking Behind the Builds Lives in the Library.

Explore Strategy LibraryContact