About us

Kestrel Health Systems builds clinical software used in 180 general practices and four private hospital groups. When our platform is down, clinicians cannot see patient records. That single fact shapes everything about how our platform team works: change is frequent but never casual, everything is auditable, and we would rather be boring than clever.

The team you would join

Platform Engineering is eight people supporting around 45 engineers across six product teams. We do not deploy other people's code for them. We build the paved road - pipelines, environments, observability, secrets, infrastructure modules - so product teams can deploy themselves safely, twenty or thirty times a week.

What you will do

• Write and review Terraform for real AWS infrastructure, starting with low-risk modules and working up
• Maintain and extend CI/CD pipelines in GitHub Actions and Argo CD
• Build small internal tools in Python or Go that remove manual steps from other engineers' days
• Investigate alerts alongside an experienced engineer, and write the incident notes afterwards
• Improve dashboards, logging and alert quality - including deleting alerts nobody acts on
• Take part in change advisory for production releases, which in our regulatory context is a real process, not a rubber stamp
• Join the secondary on-call roster from month 12, always paired, always with a senior primary
• Document everything you learn the hard way, because the next graduate will hit the same wall

Our stack

What the program gives you

• A twelve-month curriculum with named milestones rather than 'learn on the job'
• Paid AWS Solutions Architect Associate attempt in year one, Certified Kubernetes Administrator in year two
• No primary on-call in your first year. Ever. Secondary only from month 12, always paired
• Blameless incident reviews you are expected to attend and eventually write
• $2,500 learning budget and a conference of your choice each year

How the process runs

1. 1

   Application

   CV plus a short written answer: describe something you built and run yourself.

2. 2

   Technical screen

   45 minutes - Linux, networking and scripting, conversational rather than quizzed.

3. 3

   Practical exercise

   Two hours, scheduled at your convenience. Given a broken container build and a failing pipeline, get it green and explain what was wrong.

4. 4

   Systems interview

   60 minutes - infrastructure reasoning, troubleshooting under uncertainty, an incident scenario.

5. 5

   Values interview

   45 minutes - communication under pressure, judgement, how you behave when you do not know.

6. 6

   Offer

   Subject to reference checks and a National Police Check.