There are many Health+Care specific platforms now widely available and in use throughout the NHS. However, not all platforms are equal.
Our strategy at Interneuron is a platform+apps approach so the platform is fundamental to our success with almost half of our engineering team focusing on it at any one time. In fact, one of our core engineering principles is ‘Platform-first’ – if something can be centralised, shared and re-used then it should be a service or function of the platform rather than the client.
We are proud of the architecture we’ve designed and implemented so far, but we still have a long way to go. We recently released v1.1 and with several NHS deployments under our belt, we want to take a moment to share some of the thought process that has gone into designing our flagship product – Interneuron Synapse.
There are three aspects of Synapse that set it apart:
- Common and extensible data model
- Open and extensible APIs
- Insert-only data persistence
Common and extensible data model
One of the biggest benefits of a platform+apps approach is the ability to re-use data across applications – but without a structured, human readable and extensible data model its just not possible. Whilst there are some attempts to standardise Health+Care data models, we didn’t feel any of these were mature or extensible enough for our customers. Interneuron Synapse includes our Studio tool for managing, extending and building your own data models.
Open and extensible APIs
We want it to be as easy to get data out of the platform as it is to get data into it, but we also do not want to constrain how the data is accessed. We have therefore implemented:-
- Dynamic object-based APIs – These APIs allow the persistence, update and logical deleting of any object within the platform
- List and Base View APIs – These APIs allow you to create your own views of the data through the creation of SQL views which are then exposed, through our security model as APIs.
- Data Export Services – We know that sometimes you need to extract large amounts of data – for use-cases such as Business Continuity, so we enable the generation of extracts based on any List or Base View.
- Anonymisation and Pseudo-Anonymisation APIs – For research and other use-cases where identifiable datasets are not permitted.
Insert-only data persistence
We believe it’s important to be able to go back to any record at any point in time and be able to see exactly what was available at that point – therefore, we made the decision that our platform would be insert-only. The platform automatically versions, collapses and archives your data so you don’t need to worry about it. A full audit trail is available through our Open APIs too!
These are just some of design decisions we have made to date and we invite you to view our source code on GitHub – all feedback is welcomed. In a future blog post, we will take you through our microservice architecture pattern for Interneuron Synapse.