Lab
Building to Learn: What Three Weeks of Prototyping Taught Us About Agentic Marketing Optimization
A three-week prototype explored whether simulated product-intent scores could help improve AI shopping visibility. The work exposed a useful boundary: a lab can assess semantic fit, but only external evidence can test discovery or business impact.
Secure Marketing Agents: A Reference Architecture for DV360 and Google Ads Automation
How to let marketing agents analyse and propose campaign changes while identity, policy, approval, and execution controls keep authority bounded and reviewable.
The Agent Architect's Playbook: Designing Marketing Work and Reusable Skills
Define a marketing workflow, its success criteria, evidence, permissions, and evaluation plan, then package recurring expertise as a reusable skill.
The Anatomy of a Marketing Agent: From Request to Verified Outcome
How reasoning, memory, skills, tools, validation, permissions, deployment, and evaluation work together, from a single marketing agent to coordinated teams.
The Muscles of the Machine: Designing the Marketing Agent's Tool Layer
How to build and operate marketing-agent tools through MCP, CLI, or both, with explicit contracts, validation, permissions, evaluation, and deployment boundaries.
Your OpenClaw Marketing Agent: From Setup to Reliable Operation
Build a bounded OpenClaw marketing assistant, from workspace instructions and skills to authorised data access, report validation, evaluation, scheduled work, isolated deployment, and recovery.
The Code Agent's Playbook: Maintenance, Security, and Controlled Improvement
Build a coding-agent maintenance system with useful skills, independent validation, performance evaluation, enforced guardrails, isolated workers, controlled releases, and a reviewed loop for improving the harness.
Portable Intelligence: What Plugins Are and How Marketing Teams Can Use Them
A practical guide to packaging marketing skills, tools, and workflows for agent platforms, with a current comparison of Claude Code and OpenAI plugin formats, portability limits, and team governance.
The Deterministic Core: Algorithms and Data Structures for Marketing Agents
A practical guide to the search, ranking, graph, deduplication, and memory operations around marketing agents, with the assumptions and limits that matter in production.
Jev as Ad Tech's System 1: Fast Decisions, Slow Learning
How marketing agents can use Jev to classify search terms, tag creative, route leads, process customer feedback and verify proposed actions while code and people keep control.
Your Marketing Agent Needs a World, Not Another Prompt: Building Marketing Agentic Systems the Palantir Way
A marketing agent needs a reliable way to connect a request to real campaigns, approved requirements, evidence, permissions, and outcomes. Palantir's Ontology offers one useful model for that operational layer.
AI Knowledge Hub
Research projects, experiments, and resources for exploring AI in marketing.
Explore the repositoriesPractical resourceAI Skills Guide
Explore the skill catalogue and setup guidance for building marketing agents. Each resource’s status explains what is ready to use and what still needs testing.
Explore the skills guideLearning projectVisual Learning Studio
A developing approach to explaining AI and product building through visual learning.
Visit the projectAn interesting result
is the start of a question.
Experiments should explain what was tested, what was observed, and what remains unknown. Read Building to Learn for an account of the gap between simulation and evidence.
Inspect the skills guide source