Designing for Collaboration: How we rethought Ona conversations

We redesigned the way you collaborate with Ona, focusing around the relationship between you and the agent. This change creates a stronger feeling that you are working together in the same place, with the same context, toward the same outcome.

For many developers, solving a problem rarely stays in one place. You might start from a plan or a Linear ticket, then collaborate with an Agent to understand a problem, create a fix, review the changes, and eventually merge the solution. Each phase asks for something different, and when the experience is split across tools and views, your experience becomes fragmented. Context gets rebuilt instead of expanded, and it is easy to end up in a loop of "wait, what am I doing?"

The reality of shipping a feature rarely matches a clean project board. You start with a plan or a ticket, then work to find a solution, create a fix, and finally review changes. Each phase demands a different kind of focus. So, when the experience is split across disconnected tools and views, that focus becomes fragmented. You end up rebuilding context in your head instead of expanding it. It is the moment where you stare at the screen and realize you have lost the thread of the original problem.

This friction point has an impact on your productivity. Working memory is a finite resource. Every time you switch from a chat window to a code editor, you are doing more than just moving your eyes. You are forcing your brain to flush out a current set of details and reload a new one. Those small, repetitive motions, like expanding a panel to full screen just to read a function or hunting through an environment tab to find a preview, are micro-disruptions. They burn the limited mental space you need to solve your original problem.

Our redesigned experience keeps the interface calm so it only commands your attention when it's needed. Ona will hold onto the context as your goals evolve, allowing you to move through different phases of development without the friction of a context switch. You can step away for a coffee or a meeting and return to find the agent exactly where you left it. It feels less like you are talking to separate tools and more like you are working in a shared environment where the agent is in the trenches with you.

Reducing the noise

The drive for a redesign usually arrives when a product successfully grows past its original mission. We reached that point with Ona as the mission shifted from managing development environments to building a platform for background agents. To support our new direction and empower developers with agents, we had to ship fast. This daily pace allowed us to add massive value to our customers, but it also meant our interface began to battle its roots in an era that did not match our new scope. We had all the tools but needed to ensure customers were successful solving their real world problems with them.

Our conversation experience was the clearest victim of this growth. We had the main conversation focused on the left, but it was constantly competing with a right-side panel that tried to handle environments, code changes, and a full editor all at once. These are high-utility features, but they felt like they were fighting for your attention rather than supporting your work. Navigating between them started to feel like a chore instead of a shortcut. We realized that power doesn't matter if the friction of using it pulls you out of your flow.

There needed to be a major reset so you can leap forward.

Introducing a more cohesive, timeless experience

Most tools treat a chat window as a separate layer that sits on top of the actual work. For Ona, conversation is meant to start naturally and exploratory. You spend that time poking at the edges of a problem, asking questions, pivoting, and building a plan. After a plan solidifies and you find a solution, this is when you begin to narrow your focus to a single path until it is complete. Forcing these modes together without a clear information architecture creates a tool that does both poorly.

The challenge was joining these two distinct purposes without losing the utility of either. We wanted the context to be shared, the state to be visible, and the work to feel continuous rather than assembled. This is not a dashboard or a terminal. It is a workspace. By anchoring the output directly alongside the dialogue, the agent stops being a separate entity you talk to and becomes a partner looking at the same screen you are.

Today, we are revealing the result of weeks of work redesigning Ona's interface. We've rethought the sidebar, conversation view, and the density of features. These changes make space for Ona to naturally evolve with your needs, from exploration to parallel background agents. Built for high output tasks while maintaining a calm intuitive experience.

Reviewing changes

Code review is where the momentum of a task can lag. You hop out of your editor, open a PR, and rebuild the context of why a change was made in the first place. We brought the review process directly into Ona so you can decide what to do next without ever leaving the conversation.

This is also where agents trust is earned, how we make this feel empowering rather than overwhelming. To do this we structured the review around a searchable and filterable file tree and readable diffs. Keeping the logic and the code in the same view, helps you verify the impact of a change.

We are also changing the feedback loop by enabling inline comments to be sent directly back into your conversation. Making the review a meaningful part of your dialogue. You aren't just looking at a separated experience outside of your workspace; you are refining the plan in real-time. This ensures that the "why" and the "what" stay tethered together until the work is done.

Runtime and Services

The environment panel was doing too much. Ports, services, and tasks are built on top of the environment layer but they handle a different use cases. When these were in the side panel, they competed for your attention at all times. Creating friction when you are just trying to see a preview and compare to your changes.

Our new layout improves this, we moved these features into our new bottom drawer. They stay anchored and accessible without competing for the same visual priority as your dialogue or your code. In this space, they support your workflow and connect directly to the output you are driving toward.

Working with environments

In our agent-first world, the environment exists to aid a specific goal. It is the staging ground where you iterate on code, view previews, and validate solutions. Because these spaces are ephemeral, you should not have to worry about their state. This is where our old experience got in the way; we needed a redesign that moved the environment out of your face so you could focus on the outcome.

The new layout ensures the environment enables your work without demanding constant attention. You can explore changes and conversation history regardless of whether an environment is currently active. By decoupling the dialogue from the runtime state, the platform works with the context currently available. This allows you to step into a task, iterate rapidly, and move on without the friction of maintaining infrastructure.

What this unlocks

Redesigning the core experience allows Ona to expand the scope of what is possible. By moving away from a rigid environment-centric layout, we have paved the way for a product that adapts to the complexity of the task. You can solve problems across varied codebases without the friction of switching tools or rebuilding context.

This foundation also opens us up to a new set of capabilities:

• Integrated Terminals: Running a terminal directly alongside your work for immediate feedback.
• Multi-Agent Coordination: Orchestrating several agents within a single environment to tackle parallel work streams.
• Managed Sub-sessions: Isolating specific tasks without losing the high-level project state.
• Choose your agent: Bringing in external agents, like Claude Code, and running them in the same governed environment.

Use the right agent for the task, coordinated in one place.

This is a shift in how you work with Ona. We are building a system where you spend less time managing environments and more time shipping. It is about reclaiming time back to focus on the what you want to solve, while delegating the rest to agents in a way that feels safe, calm, and powerful.

You can explore the new UI in Ona today. Let us know what you think on X and LinkedIn.