FROM THE WORK
A practical starting method for clinicians who want to move into digital health without waiting for a formal AI role.
People often ask what course they should take to enter clinical AI or digital health. Learning matters, but a small piece of work in a real service can show something different: that you noticed a problem, involved the right people, made a change and checked what happened.
Choose a workflow you already understand. It might be how a practice records interpreter needs, follows up a carer, handles repeat-prescription queries or measures time to a meaningful triage response. Ask a supervisor or service lead to help narrow the question. Keep the first test small enough to manage alongside clinical work.
Write down a baseline and decide what you will measure before changing anything. Include a balancing measure: saving staff time is not an improvement if it leads to more repeat contacts or creates a barrier for patients. Speak with people who use the pathway and with the colleagues who will have to make the change stick.
Keep evidence of your own contribution: a one-page problem statement, an agreed process map, what you tested, the result and what you would do differently. You do not need patient-identifiable information in a public portfolio. If you need a concrete starting point, the Practice Lab has short guides and a synthetic-data Python exercise.
Explore the Practice Lab →