← back
Form factors for your new AI coworkers — Craig Wattrus, Flatfile
Original: Form factors for your new AI coworkers — Craig Wattrus, Flatfile
Takeaway
AI form factors range from invisible to conversational, and the designer's new job is character-coaching the model and crafting the right box around an inherently parallel collaborator.
Summary
- Flatfile classifies agent surfaces into four buckets: invisible (background app generation from email signup), ambient (sparkles surfacing data issues), inline (in-row data fixes via code execution), and conversational (build mode)
- Reframes agent prompting as 'character coaching' (inspired by Amanda Askell on Claude's character) rather than copy editing
- Built a V0 'chat tuner' to feel out how a Claude system prompt responds to traits like friendly/balanced/concise
- Themes: feel the material (work with the LLM directly), find the grain (where it's strong vs brittle), and design boxes that let the 'PhD intern' shine
productuiagents
Original description
Designing user experiences for AI means moving beyond traditional interfaces. Designers are grappling with how to create intuitive and effective interactions for these new AI capabilities, while growing their practice to include philosophy, ethics and coding. What if AI interactions could be reimagined as new 'coworkers'? This talk explores AI systems as your new coworkers. Covering novel UX patterns we’ve implemented and are researching at Flatfile as well as a state of the union on emergent patterns we’re seeing and using from the industry. Attendees will get a peek into explorations into AI cursors, forward-leaning chat paradigms and tool UX. We will discuss both work thats in production today at some of our biggest customers as well as thought-provoking demos, offering a vision for the future of AI UX. Timestamps 00:25 Design Engineering: Form Factors for your new AI coworkers 01:04 Four main categories of AI stack: invisible, ambient, inline, and conversational 02:14 Invisible AI: Personalized demos 03:06 Ambient AI: Analyzing data for opportunities 03:10 Inline AI: Direct data manipulation 03:46 Shifting from helicopter parent to character coach 04:32 The "chat tuner" tool 05:20 Feeling the material of AI 08:20 Finding the grain in AI design 11:08 Courting emergence 11:48 Example of emergence: Combining datasets and generating reports 12:37 Example of emergence: Suggesting human intervention 14:09 Eyes on the future 14:27 "Auto-complete" for data transformations --- Craig Wattrus AI Design Engineer Craig Wattrus is a product designer and technologist working at the edge of human-computer interaction and AI. He designs and codes at Flatfile, where he’s leading a product called AI Transform. He's building adaptive data systems that use agentic AI to automate complex workflows across Fortune 500 companies. With a background in both computer science and design, Craig’s work focuses on shaping new UX patterns for AI systems that observe, learn, and act alongside users not just for them. Craig is deeply interested in rethinking form factors for AI, exploring how designers can create new patterns of interaction that feel more collaborative, contextual, and adaptive. His approach blends practical, production-ready work with speculative design exploration with working PoC's offering a grounded yet forward-looking take on what AI-native UX can become. When he’s not dreaming up new ideas or testing AI behaviors in production, he’s probably making lamps, tinkering with small-scale hardware, or enjoying a perfect espresso.