Forecasting & planning
Sales, demand, capacity and cash-flow forecasts with uncertainty ranges, updated automatically as new data arrives.
Consultants based in Stockholm · on site or remote
Data Science & Advanced Analytics
We turn your data into answers: forecasts you can plan with, experiments that prove what works, and models that run in production, not just in a notebook.
Sound familiar?
- Forecasts are built by hand in spreadsheets, and nobody fully trusts them
- Decisions are made on gut feeling because analysis takes too long
- A model was built once but never made it into daily use
What we deliver
How we can help
Experiments & A/B testing
Test design, sample sizes and statistically sound analysis, so you know which change actually made the difference.
Customer analytics
Segmentation, churn prediction and customer lifetime value, so effort goes to the customers that matter most.
Pricing & optimisation
Price elasticity, assortment and resource optimisation, built on your own data and business constraints.
Exploratory analysis
A fast, structured look at your data to find patterns, outliers and the questions worth answering.
Models in production
Scheduled scoring, monitoring and retraining, with results delivered to Power BI, apps or your business systems.
How we work
From idea to result
- 1
Understand the decision
We start with the decision the analysis must support, who makes it and how often.
- 2
Explore & prototype
We explore the data and build a first model quickly, so you can judge the value early.
- 3
Validate
We test the model against history and simple baselines, and explain what drives its results.
- 4
Put into production
The model runs on schedule, is monitored, and its output appears where people already work.
Typical assignments
Where this makes a difference
Demand forecasting
Weekly forecasts per product and location that replace a manual spreadsheet process and feed purchasing and staffing plans.
Churn prevention
A risk score per customer, updated every night, so the sales team knows whom to call first.
Measuring what works
Controlled experiments on prices, campaigns or website changes, with clear answers instead of opinions.
Technologies we use
- Python
- pandas
- scikit-learn
- statsmodels
- Prophet
- PyMC
- MLflow
- Databricks
- Power BI
- Streamlit
Frequently asked questions
Do we need a data platform first?
No. We can start with the data you already have, such as exports, databases or Excel, and build the platform step by step if the value is proven.
How quickly will we see results?
We aim to show a first working prototype early, so you can judge the value before investing further.
Who owns the models?
You do. Code, models and documentation are delivered to you, and we make sure your team understands how they work.
Let's talk data.
Tell us where you are today and where you want to go. We will help you find the shortest path there.