Ableton Live
Exploring adaptive interfaces in complex software
Overview
For this project I explored whether adaptive interfaces could make complex creative tools easier to use, without taking away the control that makes them worth using in the first place. Using Ableton Live as a case study, I researched adaptive interface frameworks, explored the tool as a first time user, and observed a "hobby" producer workflow to identify where adaptation could meaningfully reduce friction.
The outcome is a speculative Figma prototype demonstrating specific adaptive behaviours within Ableton's existing interface.
The problem
Every producer uses Ableton differently. However, the interface is the same for everyone.
Can adaptivity make a complex tool easier to use, without taking away the control that makes it worth using in the first place?
Opportunity
GenAI is making adaptive interfaces viable for the first time.
As the barrier to building and shipping software gets lower, products are increasingly launched with adaptable foundations and refined through real use, rather than being fully optimised from day one.
This creates a real opportunity for complex tools like Ableton: the interface no longer needs to be a fixed structure that every user adapts to. It can be a system that learns from how each person works and reduces friction over time.Solution
Rather than redesigning Ableton, the focus was on designing adaptive elements that could sit within the tool as it is, that would reduce friction at specific moments without disrupting the workflows producers already rely on.
The research identified three recurring moments where adaptive behaviour could meaningfully reduce effort. These weren't designed as isolated features but as patterns that kept showing up across the observed workflow.
Pattern recognition in recording
Producers repeatedly identify and manually group related elements that consistently work together. Visual grouping or colour-coding based on those patterns could make that automatic.
Configuring macros for filters
Adjusting multiple filters means constant back-and-forth between them. Coordinated editing across similar filters could avoid that repetition without changing how the underlying workflow works.
Overview of related MIDI clips
Comparing or editing related MIDI clips across different tracks requires switching between them manually. Better cross-track visibility would support that without restructuring anything.
Of the three, the filter macro configuration was developed into an initial design. The other two remain as ideas on paper.
Core flows
Research
The research was structured around four stages, each building on the previous one.
Before touching Ableton, I needed to understand the space. That meant reading research papers on adaptive interfaces, exploring existing projects, and mapping out available frameworks. Two things shaped the rest of the project from this stage.
The first was understanding what adaptive interfaces actually are. The second was the Norcio and Stanley (1989) framework, which defines four knowledge domains that any adaptive system needs to reason about: knowledge of the user, the task, the interaction, and the system itself. That framework became the backbone for how I structured the rest of the research.
Because this project sits at the intersection of AI and creative tools, I wanted to understand how AI is already being used in music production and where producers feel it belongs. I read research papers on the topic but also spent time reading personal blogs and interviews from working producers, which gave me a much more honest picture of how people in the field actually feel about it.
Producers are open to AI as something that supports and accelerates their workflow, but strongly resistant to it acting as a creative agent. AI that handles repetitive, technical, or organisational tasks is welcomed.
That distinction became a guiding principle for the design: adaptation should reduce friction in the workflow, not participate in the creative process.
Coming in with zero Ableton experience, I worked through their online beginner guide, downloaded the tool, and watched tutorials to learn to navigate it. Three things stood out: the interface is overwhelming from day one, the built-in Info View helps with individual elements but doesn't help a new user understand how to start a project, and learning depends almost entirely on external resources rather than anything in the product itself.
To understand how producers actually work, I reviewed online discussions to identify recurring pain points, then built a NotebookLM repository combining YouTube workflow videos and the Ableton Live handbook. Using the language model, I generated multiple user journeys representing different workflow types from beginner to advanced. These became hypotheses to pressure-test against a real user.
A combined observation and ideation session with a real Ableton user. They worked through an existing project, explaining their decisions as they went. We discussed the workflow in real time and reflected together on moments where adaptation could reduce effort, which turned out to be mostly about personal preferences and recurring patterns specific to their way of working. Micro-friction moments were captured as sticky notes and prioritised by potential impact, leading to three adaptation opportunities.