Exp. 06
Vega & Enso
A personal research thread in embodied AI — the spatial agent, then the product
- Role
- Personal research — design & build
- Year
- 2024—
- Stack
- Eleven Labs + OpenAI APIs, app design
Between client work I run a personal research thread on embodied AI — what makes an artificial companion feel alive, and what it’s actually for. Vega came first, a spatial agent; Enso is its evolution, the same questions reshaped into a product. Both are living works in progress.
Vega — the spatial agent
Vega is a spatial companion prototype — an exploration of how spatial agents look, move, and build a relationship with you, spanning 3D design, technical art, and product design.
I chose particles as its form: a liquid, adaptive shape that reshapes with the conversation and gives the agent clear ways to express listening, thinking, and talking. Voice and reasoning run on Eleven Labs and OpenAI.
Enso — from agent to product
Enso carries Vega’s questions into a product — an embodied AI companion for managing your goals and focuses, visualised as fluid, interactive bubbles rather than fixed lists and folders. It’s aware of everything in the app, learns from your conversations, and can act on what you tell it.
The form matters as much as the function: Enso should feel alive without tipping into the uncanny valley, so I deliberately avoid mimicking human behaviour. It’s an early prototype — the core features are built out, with a first round of user testing done.
A recorded session with Enso — moving through goals and focuses in conversation.
Form study — how Enso's shape shifts as it moves through its states.
Ears surface on the character when Enso is listening.
The project wound down after I decided to join Tolan full time. I’d had plans to keep developing it, and may yet return to it.
Credits
- Everything — Rafe Johnson