Right, let's be upfront about a few things. We're a 30-person startup in Sydney building predictive analytics for the hospitality sector — think demand forecasting for restaurants, staffing models for hotels, waste reduction for catering companies. The data is sometimes weird. The clients don't always give us clean exports. Some of our dashboards still live in Excel because the client's ops manager refuses to change. That's the reality, and we're not going to pretend otherwise in a job ad. What we are is genuinely good at what we do, growing fast, and looking for a junior data scientist who actually enjoys the messy early stages of an analytics problem — cleaning data, asking dumb questions that turn out to be important, and building something simple that actually gets used. You'll work closely with two senior data scientists who are invested in your growth. We do regular code reviews, we pair program a few hours a week, and we celebrate wins properly. If that sounds like a good place to learn, we'd like to hear from you.
Responsibilities
Clean, explore, and validate data from client systems including POS exports, booking platforms, and spreadsheets
Build and evaluate predictive models for demand forecasting with senior guidance
Create clear visualisations and data summaries for client-facing reports
Document your analysis so the team can reproduce it and build on it
Participate in sprint planning and weekly retrospectives
Requirements
Solid Python for data work — Pandas, NumPy, Matplotlib at a minimum
SQL you can actually write from scratch: joins, aggregations, window functions
Basic statistics you understand rather than just apply — confidence intervals, correlation vs causation, overfitting
Scikit-learn for model training and evaluation
Curiosity about messy real-world data and patience when things don't go to plan
Any personal project, university capstone, or bootcamp work involving real data is very welcome