Research
We study new interaction models and interface patterns for emerging technologies.
- Human–AI interaction
- Agent interfaces
- Generative UI
- Human-in-the-loop systems
- Adaptive interfaces
- Delegation + approval
Interfaces between people and technology
We research, prototype, and build better interfaces between people and technology.
We explore how emerging technologies—especially AI—change the way people interact with software, and turn those ideas into working products.
01 / Definition
What is a good interface?
A good interface makes complex technology understandable without hiding what matters.
It gives people the right information, control, and feedback at the right moment.
As software becomes more intelligent and autonomous, designing the boundary between people and technology becomes increasingly important.
02 / Practice
What we do
Prototype
We turn ideas and hypotheses into working software quickly enough to test them in the real world.
- Prototype to think
- Prototype to test
- Prototype to learn
Build
We design and build production-ready interfaces and software based on what we learn.
- Idea
- Prototype
- Test
- Learn
- Build
- Ship
03 / Current focus
Interfaces for AI
AI changes software from a deterministic tool into something that can interpret, generate, decide, and act.
We are exploring how interfaces should evolve when software becomes less predictable and more autonomous.
04 / Experiments
Planned investigations
Research directions we want to test through working software. These are planned—not published findings.
01Approval interfaces for autonomous agentsPlanned
Exploring when an agent should ask, suggest, or act.
- Type
- Experiment
- Status
- Planned
- Topic
- Human control
- Question
- How can approval feel consequential without becoming a constant interruption?
02Beyond the chat boxPlanned
Exploring alternatives to conversational interfaces for structured AI workflows.
- Type
- Experiment
- Status
- Planned
- Topic
- Interaction model
- Question
- Which kinds of work benefit from direct structure instead of conversational turns?
03Interfaces for uncertaintyPlanned
How should software communicate that an AI may be wrong?
- Type
- Experiment
- Status
- Planned
- Topic
- AI feedback
- Question
- What information helps a person calibrate trust and choose the next action?
04Generative interfacesPlanned
Exploring interfaces whose structure changes in response to user intent and system state.
- Type
- Experiment
- Status
- Planned
- Topic
- Generative UI
- Question
- How can an interface adapt while remaining coherent, learnable, and controllable?
05 / Principles
How we work
01Make complexity understandable.
Reveal the model people need without flattening the parts that affect their choices.
02Give people meaningful control.
Put intervention, approval, and recovery where they change an outcome.
03Prototype before assuming.
Use working systems to test the interaction, not just the presentation of an idea.
04Design systems, not screens.
Consider behavior, state, feedback, and failure as one connected interface.
05Build what we design.
Carry intent through implementation, where the real interface decisions are made.
06 / About
Good Interface Lab
Good Interface Lab is an independent design and engineering practice focused on interfaces between people and technology.
We combine interface design, software engineering, and emerging technologies to explore new ways of interacting with increasingly capable systems.
Work with us
We work with teams exploring new products, interactions, and workflows—especially where existing interface patterns are no longer enough.
- Interface exploration
- Product prototyping
- AI-native product design
- Product design + engineering
07 / Contact
Start a conversation
Have a difficult interface problem? Let’s look at it.
hello@goodinterface.dev