How We Work: Building Shared AI Workflows

Most research teams are in the same place right now: individuals are experimenting with AI on their own, but there is no shared process, and growing pressure from leadership to move faster.

The result is uneven output quality, duplicated effort, and a creeping skepticism about whether AI is actually helping.

What teams tell us they actually need isn't another AI tool or a prompt guide. It's clearer direction on when AI is worth using, and a shared understanding of how everyone on the team will use AI -- not just encouragement to experiment more.

In How We Work, your team will:

• Know exactly where AI fits in your research workflow — and where it doesn't — so individuals stop making that call alone

• Use AI to cut 10+ hours from synthesis on tightly scoped studies without losing the rigor stakeholders expect

• Develop deeper insights by applying AI as a thought partner during complex sensemaking: stress-testing insights, surfacing alternative interpretations, and unblocking synthesis when the team is stuck

• Share AI-supported outputs with stakeholders confidently, with a clear process to show and defend

• Walk away with shared norms, reusable templates, and quality standards the whole team operates from

Pre-work — Each team member completes a short survey on their current AI use and confidence. You get a workflow audit that maps where AI is already showing up, where the gaps and friction points are, and what the team needs to get out of the program.

Workshop 1: Current state analysis — You review the audit together, align on shared goals, and identify the highest-priority workflow problems to solve. The team learns the core skill of thinking with AI through hands-on practice, and defines 3–4 focused experiments to run during the practice week.

Practice weeks — Team members apply selected AI methods to real project work, with guided exercises and individual coaching support. You document what worked, what didn't, and what was missing and what you'd recommend implementing.

Workshop 2: Learning and alignment — You debrief the experiments, align on what methods and standards to carry forward, and define your team's guardrails and enablers. You leave with a refined future-state workflow, a shared language for AI use, and agreed next steps for adoption.

*This is a learning framework that's entirely customizable based on the team's needs and process. Most teams choose to add skill development modules, more coaching weeks or implementation support.

I truly enjoyed this class and learned so much from it. Seeing how Ann thoughtfully approaches when, why, and how to work with AI was especially illuminating. Her process has become a real model for how I now incorporate AI into my own work-- more intentionally, strategically, and creatively. The resources she shared were also incredibly valuable. I have returned to the LLM Playbook and other course materials again and again, and they have become go-to references for me. Thank you so much for such an engaging, practical, and genuinely helpful workshop!

• Maaike B., Clinical Associate Professor · NYU

Excellent course. If you're looking for ways to thoughtfully integrate AI into your design practice, this is the course for you. I'm a learning designer, and really appreciated the thoughtful way the course was organized to provide an introduction to and then opportunities to practice skills in really concrete, relevant ways. Ann brings a refreshing combination of depth of knowledge and experience, and a continual openness and curiosity to learning new things and adjusting course content as AI technology continues to advance.

• Aaron, Program Lead, Civilla