FlexAPI
Personalised behaviour change content, generated through an API.
FlexAPI is Onmi’s backend for generating personalised behaviour change programs. It combines AI with the Do Something Different methodology to create Do’s: small, practical actions that encourage people to explore behaviours beyond their usual routines.
FlexAPI provides the content generation layer. Applications and services provide the user experience. FlexHabits, Onmi’s behaviour change app, is one front-end built on FlexAPI.
What FlexAPI generates
FlexAPI generates personalised Do’s within behaviour change programs. Each Do translates the approach into a concrete action that someone can try in everyday life.
The focus is on behavioural experimentation: encouraging people to act differently in familiar situations and expand the range of responses available to them.
A Do might invite someone to initiate a conversation they would normally avoid, change part of a familiar routine, or approach an everyday decision differently.
How the methodology shapes the content
FlexAPI is grounded in Do Something Different, developed by Professors Ben (C) Fletcher and Karen Pine.
The approach focuses on behavioural flexibility: the capacity to use a wider range of behaviours as circumstances change. Rather than relying solely on information or advice, it encourages people to practise unfamiliar behaviours through small, manageable actions.
This principle shapes the content FlexAPI generates. Do’s are intended to be actionable experiments, giving users an opportunity to experience a different way of behaving.
Do Something Different model of behaviour change. New actions lead to new cognitions. There is no cognitive route from old habit to new action. [1]
Why a Do may look beyond the target behaviour
A program’s goal provides direction, but its Do’s do not necessarily address that goal directly.
Everyday behaviours occur within connected routines. A particular setting, journey, interaction, or sequence of actions can cue what happens next. The Do Something Different approach therefore considers the wider behavioural patterns surrounding a target behaviour.
For example, a familiar journey home might lead someone past a shop, cue a purchase, and become part of a repeated evening routine. Trying a different journey introduces a change earlier in that sequence.
This explains why a program focused on a particular goal may include Do’s about social interactions, travel, leisure, or other everyday activities. The intention is to create opportunities for different responses across the routines that surround the goal.
Personalisation and context
FlexAPI can draw on contextual data to personalise generated content.
Context can help shape which actions are relevant, how they are expressed, and when they are appropriate. This supports adaptation to the user's circumstances and the setting in which a program is delivered.
Personalisation connects the methodology to everyday opportunities for action.
Separate generation from delivery
FlexAPI and FlexHabits serve different roles:
FlexAPI Generates personalised behaviour change content.
FlexHabits Provides an app experience through which users engage with that content.
This separation allows content generation to sit behind a user-facing application, while the application manages how people encounter and interact with their program.
Research foundations and ongoing development
The Do Something Different methodology emerged from the research of Professors Fletcher and Pine into behaviour and behavioural flexibility.
FlexAPI is Onmi’s implementation of this approach, bringing its principles into an AI-supported content generation system.
Professor Fletcher continues to collaborate with Onmi on the development of our behaviour change programs.
Integrating FlexAPI
FlexAPI provides the content generation backend for user-facing applications, including Onmi’s FlexHabits app.
For details on working with the API, see our API documentation.