Data science and forecasting, on contract
Forecasting, machine learning and predictive models, built on contract to support existing decisions. Suited to questions about what is likely to happen next, rather than what has already happened.
Spreadsheets describe what has already happened. They are of less help in estimating what comes next: demand next month, a water level tomorrow, or which bookings are likely to cancel. Where there is sufficient history, a forecasting or predictive model can fill that gap.
What we build
- Forecasting models, for demand, levels, bookings and anything else measured over time
- Predictive and classification models, to anticipate outcomes in advance
- Statistical analysis, to test whether a difference is significant
- Models delivered through a dashboard or an API, so they are used day to day
Typical use cases
- Forecasting demand to plan staffing and rosters
- Predicting stock requirements to reduce over-ordering
- Estimating which bookings or orders are likely to cancel
- Identifying customers who are at risk of leaving
- Flagging unusual transactions for review
- Forecasting cash flow over the coming months
Decision support
A model gives decision makers a clearer view of what is likely, and the decision remains theirs. Every forecast is reported with its typical error. A model is only as reliable as the history it is trained on, so it is least reliable in conditions it has not encountered. We set out those limits at the start of each project.
Tools we specialise in
- Python, with pandas and scikit-learn
- SQL
- Cloud hosting
Business Health Check
An AI Readiness Check, a Process Review and a Reporting Setup, at fixed prices.
From $450 + GST, or all three for $1,500 + GST
Talk to us about your business
An initial conversation is free and carries no obligation.