Automattic's AI Fluency Strategy

Automattic, known for supporting platforms such as WordPress.com and Tumblr, operates as a fully distributed company with over 1,400 people across 82 countries.

The company's bet is that employees need real time to learn and experiment with AI tools to genuinely integrate them into their workflows and develop meaningful skills.

Automattic developed two-week intensive, in-person enablement programs for its employees. These immersive experiences are structured to balance facilitated learning with practical application.

The curriculum consists of 50% workshops and 50% real project work, encouraging participants to daily showcase their workflows and project demos while developing tools, dashboards, or Model Context Protocols (MCPs) for daily tasks.

They bring people in person for a two-week immersive experience, emphasizing hands-on learning.

This customization ensures relevance for various roles within the company.

The engineering curriculum covers foundational AI concepts, creating Model Context Protocols (MCPs), spec-driven development, and responsible AI usage, while also touching on advanced topics like hook systems, autonomous agents, and applied AI features for WordPress. The rapid evolution of AI models necessitates constant curriculum updates to ensure the content remains current and effective.

Automattic developed 'context-automattic', a custom internal Model Context Protocol (MCP) server, to manage governance and observability for its AI initiatives. This infrastructure supports the goal of scaling AI expertise across the organization.

The company aims for each program participant to become a 'multiplier' within the company, extending AI knowledge and mentorship to colleagues who have not yet attended the formal training. This strategy helps reach more employees beyond the initial cohort.

This year, Automattic completed seven cohorts, aiming to train 600 total attendees, approximately two-thirds of its engineering team. They also facilitate regional AI meetups to sustain knowledge sharing and build community among distributed teams.

This method is more effective for skill development than abstract conceptual discussions.

The company also built a shared skill and Model Context Protocol (MCP) infrastructure to facilitate collaborative development and maintain an 'always-on' Slack channel, offering support from program graduates to those new to AI workflows.

Automattic tracks the success of these programs by assessing 'intentionality'—the transition toward agentic workflows—and 'flow' to verify how well AI integrates into daily operations. Knowledge sharing is also a priority, aiming to transform every employee into an AI teacher and expert.

Surveys conducted at 72 hours, 30 days, and 90 days post-program show that participants continue to gain confidence and experience, with sustained growth in their ability to apply and experiment with AI tools over time.

During two-week cohorts, employees produced tangible assets like a WooCommerce analytics agent and a Zendesk connector that has since been adopted for company-wide support. These projects demonstrate immediate application of learned skills.

A Zendesk connector was also created to enhance support operations, and this tool has since been adopted across multiple company teams, improving workflow for customer service. Participants also built an internal 'skills directory' to support diverse AI tools like Codeium and Claude, avoiding reliance on a single vendor.

Notably, a support employee, who felt more comfortable with AI after completing one of the programs, developed a guided admin tool to assist users during help sessions by using AI to highlight screen navigation.

The core lesson is that meaningful change in working methods requires more than just introducing new technology.

Education programs must foster a safe environment where learning is the primary focus. Transforming how staff work effectively necessitates creating an environment conducive to change rather than simply expecting new tools to drive it.