Welcome to InfoSum!
We specialize in the non-movement of data.
Our Platform provides the infrastructure that enables companies to collaborate with data without any of the privacy or commercial trust risks that exist today. We differ from other solutions as there is no centralisation or sharing of data. Advertisers, publishers and data owners can work together to overlay their first-party data without sharing it. They can understand the correlation between their audiences and develop strategic partnerships built on transparency and the safe isolation of data.
Each company’s data, once uploaded to a Bunker and normalized to a standardized schema, never gets shared with anyone and only the dataset owner is able to access it. When a user uploads several datasets, each to separate Bunkers, the insights are automatically ‘connected’ via a mathematical model, but the data never actually goes anywhere. Instead, the user can see statistical information on the relationship between the data, including quality metrics, the overlap of identifiers and the types of attributes held within the data.
Likewise, users can share permissions with others to connect their data and discover their affinity. This facilitates an entirely new approach to data collaboration, as multiple parties can compare and contrast their data without being able to see each other’s data or hand over anything. Permissions can be turned on and off or changed at any time, so each party is always in total control of their data. New partners can also be discovered within the Platform to assist with exploring, enriching and segmenting data.
Sophisticated insights can be generated by further analyzing the connected data. This can be done by comparing several sources at once, running a series of aggregations and filtering the data on a particular attribute or trait. Alongside the non-movement of data, differential privacy concepts are used throughout to make it impossible for anyone to misuse the system to re-identify an individual. These insights can be activated against by using an activation dataset, provided an additional permission is granted by each data owner involved in the analysis.
In short, using the Platform, you can:
- Onboard and standardize disparate data
- Match individuals using keys and 'fuzzy' data
- Use attribute data to segment audiences
- Discover enrichment data sources
- Visualize and activate query results
This site provides a rich library of concepts, tutorials and reference material. You can use the search bar or the menu to explore the content. For a quick tour of the Platform, please see the using tutorial mode article.
To get started with importing a dataset, please see our quick start guides here:
- Creating a dataset
- Accessing a Bunker
- Importing your data
- Normalizing your data
- Publishing your data
- Viewing the connectivity
- Writing your first query
- Sending a permission
It is worthwhile taking the time to read the following reference material:
- Data processing concepts
- Privacy controls
- Protecting personal data
- Quality metrics
- Data processing concepts