Services
Data Mining
Finding the structure in data that nobody thought to look for.
Overview
Data mining uses algorithms to discover patterns, groupings, associations and outliers in large datasets. It is particularly useful when the questions are not yet well formed and the goal is to learn what the data contains.
We pair automated discovery with domain review, since many discovered patterns are artefacts of how data was collected. Patterns that survive scrutiny are turned into hypotheses, features or rules.
Problems we help with
- Large data stores whose contents are poorly understood
- Segments and behaviours not visible in standard reports
- Suspected fraud or error without a clear signature
- Text and log data that nobody reads at scale
Capabilities
Clustering and segmentation
Grouping customers, products or events by behaviour.
Association and sequence mining
Frequent patterns and next-step relationships.
Outlier and anomaly discovery
Statistical and model-based detection of unusual cases.
Text mining
Topic discovery, entity extraction and sentiment from unstructured text.
Process mining
Reconstructing real workflows from event logs.
Pattern validation
Significance testing and holdout checks to separate signal from chance.
How we approach it
- 01
Profile
Understand structure, coverage and bias in the data.
- 02
Explore
Run discovery techniques with multiple settings.
- 03
Validate
Test patterns against fresh data and domain knowledge.
- 04
Operationalise
Convert accepted patterns into features, rules or monitoring.
Where it is applied
- Customer behaviour segmentation
- Basket and cross-sell analysis
- Detection of unusual transaction sequences
- Mining service tickets for recurring causes
Outcomes to aim for
These are goals we design towards, not guarantees.
- A documented map of what the data contains
- Validated segments and patterns
- New features for predictive models
- Targeted anomaly review lists
Frequently asked questions
Bring us a decision you want to improve.
Tell us about the problem and the data you have. We will reply with a candid view of where we can help.