Interaction Designer

How do you design an interface when the product can act on its own?

A1 is building an AI assistant designed to do more than respond to prompts. The goal is to create a proactive system that can maintain context, work through multi-step tasks, interact with external tools, and help users complete everyday work with significantly less manual effort.

That creates an unusual interaction design problem: how do you give AI more autonomy without making the person using it feel like they've lost control?

A1 is hiring an Interaction Designer to work specifically on that challenge.

What you'll design

This role goes deep into the mechanics of human-AI interaction rather than focusing primarily on visual styling.

You'll define what happens while an AI system is working: what the user sees, what they can change, when the system should ask for permission, and what happens when something goes wrong.

Your work will include:

• Designing flows for starting, monitoring, pausing, correcting, and completing AI-driven tasks.
• Defining states such as thinking, acting, waiting, uncertainty, interruption, failure, and recovery.
• Creating ways for users to review and override AI decisions without having to supervise every step.
• Designing patterns around delegation, approvals, corrections, escalation, and reversibility.
• Working through edge cases where AI behavior may be unpredictable.
• Building detailed prototypes and interaction specifications for engineering.
• Testing concepts with users to understand where AI behavior creates confusion or reduces trust.
• Developing reusable HCI patterns that can eventually be applied throughout the product.

The interesting part: there isn't a mature playbook yet

Traditional software usually behaves predictably. An action produces an expected result.

AI systems don't always work that way.

A1 wants someone comfortable designing around non-deterministic behavior, where the same system may behave differently depending on context and where errors need to be anticipated rather than treated as exceptional cases.

You'll need to make decisions such as:

Should the AI ask before doing this?

Does the user need to see what it's doing?

Can the action be easily reversed?

When should the system interrupt the user?

How does it explain uncertainty without becoming annoying?

Those decisions are a central part of the job.

What they're looking for

A1 is interested in Interaction Designers or experienced UX Designers with particularly strong interaction-design skills.

You should be comfortable working at a detailed level with states, transitions, flows, edge cases, feedback, and failure scenarios, while still understanding how those decisions fit into the broader system.

Strong prototyping skills in Figma or comparable tools are expected.

You'll also need a solid understanding of HCI principles around user control, transparency, feedback, trust, and recovery, together with experience collaborating closely with engineers.

Systematic thinking will matter more than simply producing polished screens.

How the team works

A1 describes itself as a small, hands-on team that moves quickly and expects people to operate with substantial independence.

For design, that means being able to bring structure to problems where established interaction patterns may not exist yet, make decisions with incomplete information, test them, and iterate quickly.

Candidates will need to submit a portfolio. The hiring process includes two interview rounds and a design assessment.

For an Interaction Designer interested in human-AI interaction rather than simply adding AI features to conventional interfaces, this is the kind of problem space where many of the UX patterns are still waiting to be invented.