FreezePoints eMagazine

On the evening of Tuesday, April 14, 2026, PMI Orange County welcomed Jen García to Pali Wine Co in Anaheim in anticipation of her sharing her experience and expertise. García, AI Strategy and Partnerships Lead at Caltech's Center for Technology and Management Education (CTME) and Founder and CEO of Ingrained, delivered a presentation entitled "Beyond the Hype: How an AI Strategist Uses Claude for Project Management and Why LLM Communication Skills Beat Tools."

Her central thesis was both refreshing and practical: the tool matters far less than your ability to communicate with it. In an era of rapid AI adoption, García argued that learning how to effectively engage large language models (LLMs) is quickly becoming a core professional competency — one that project managers are uniquely positioned to master.

Communicating Smarter, Not Just Faster

García assessed attendees’s fluency of advanced prompting. She asked about elements like establishing the LLM's role, defining your audience, specifying format, and asking the model to explain its reasoning all contribute to more reliable, useful results. She also introduced two evolving disciplines that she expects to grow in importance: context engineering (selectively curating the files, documents, and background an AI needs to perform a task) and intent engineering (going beyond policies to provide the AI with your organization's mission, values, and strategic goals). She shared a Klarna story to exemplify what happens when AI follows the letter of policy while missing its spirit entirely.

She also referenced a notable MIT study, "Your Brain on ChatGPT," which found that users who engaged their own thinking before turning to AI produced the best outcomes — a finding that affirms rather than undermines the role of human judgment in AI-assisted work.

Agents, Tools, and What's Coming

García gave the group a grounding in agentic AI — the emerging paradigm in which an LLM is paired with data and tools (such as web access or software integrations) to autonomously complete multi-step tasks. She emphasized that context and intent engineering become even more critical as organizations begin deploying AI agents, since those agents will only be as purposeful as the instructions and values built into them.

She also highlighted a pair of practical tools worth exploring: Granola for AI-assisted meeting notes, and Wispr Flow for HIPAA-compliant voice-to-text transcription.

Your Homework

García left attendees with a clear challenge: start with Claude Projects — contained conversation threads with persistent instructions and memory — and then level up to Skills, which function as reusable, deeply instructed templates for recurring tasks like status reports, agendas, and presentations. Both are available today and require no technical background to get started.

For those looking to go deeper, Caltech CTME offers a growing catalog of AI and project management courses — many carrying PDU/CEU credit — including AI for Project Management and the hands-on Claude Code Agent Lab. Visit ctme.caltech.edu to explore the full course lineup.

______

This article was co-written by Roderick Conwi and Claude (Anthropic), an AI large language model. The content was developed using notes and materials from the April 14, 2026 PMI Orange County Dinner Meeting, including the presenter's Granola.ai session notes and slide deck. All editorial decisions, contextual framing, and final review were made by the human author.