Data Scientist

About the role

CloudPeak has been running other people's infrastructure since 2015, which means we are sitting on six years of change records, incident timelines, capacity curves and support transcripts across forty-two customer environments. It is modelled reasonably well in the warehouse. It has almost never been asked a hard question.

The Data & Insights team is five people — two analytics engineers, an analyst and a data platform engineer — and this is the first dedicated data scientist role. You will report to the Head of Data and work most closely with the reliability team, whose leaders want to know which changes actually predict an incident, and with the commercial team, who would like to stop guessing which accounts are about to grow.

The honest version: half of this job is analysis and modelling, and half is being the person who says out loud that a column has meant three different things since 2022. Both halves are respected here.

What you'll do

• Build and validate models that matter to the business: incident likelihood after a change, capacity headroom, account expansion and churn risk.
• Own analyses end to end — question, data, model, write-up, and the conversation with the team that has to act on it.
• Extend the dbt models in the warehouse so your work is reproducible by someone else next quarter.
• Design and read experiments on platform and console changes, including deciding when an experiment is not worth running.
• Publish findings that survive scrutiny from engineers who know the systems better than the data does.
• Improve data quality at source by working with the platform teams on what gets instrumented and how.

What we're looking for

• Four or more years in applied data science, with production work you can describe in detail.
• Strong SQL and Python — pandas, scikit-learn, and enough statistics to know which assumptions you are breaking.
• Experience with a modern warehouse stack (dbt, Snowflake, BigQuery or equivalent).
• The ability to explain a model's limitations to a sceptical infrastructure engineer without hiding behind jargon.
• Business-level English; the data team works in English.

Nice to have

• Time-series or survival analysis on operational data.
• Familiarity with cloud infrastructure concepts — you will be modelling them.
• Experience presenting to customers as well as to internal teams.

What you get

• 13th month salary plus discretionary annual bonus
• Medical and dental for you and your dependants
• MPF employer top-up to 9%
• 20 days annual leave, rising to 25
• HK$20,000 conference budget — one overseas conference a year
• Shuttle bus from University station