Getting Started: Developer Guide for Endeavour Predict tools
Endeavour Predict provides Software Development Kits for integrating health‑risk prediction models into your applications, as well as online tools such as APIs and the Batch Processor which do not require installation.

This Guide signposts developers to the essential information for accessing the platform, using the tools, and embedding software components.

1. Accessing the Portal

Developers should begin by signing in. To create an account contact us.

Two account types are available:

  • Basic Accounts for users who only require the API and/or the Online Batch Processor
  • SDK Accounts, for system integrators who want to work with the DLLs

SDK, API and Online Batch Processors all require a key stored in EP Engines → My Keys.

Keys are assigned to your Organisation by Endeavour Predict when we create your account.

2. Understanding the Platform Architecture

Endeavour Predict is available through four paths:

2.1. Web APIs

Stateless endpoints for real‑time risk scoring. Ideal for:

  • Cloud applications
  • Web and mobile apps
  • Cross-platform integrations
2.2. Software Development Kit (.NET)

Local components for environments where:

  • On‑premise execution is required
  • Data cannot leave the organisation
  • Low‑latency, offline, or embedded execution is needed
2.3. Online Batch Processor

A tool for large‑scale data processing of Comma Separated Data (CSV). Useful for:

  • Research teams
  • Non-technical audiences
  • Population health analytics
2.4. Web Calculator

A freely available tool for single patient Risk scores, useful for:

  • Exploring personalised health risks using validated clinical models.
  • Clinics, research settings, and public health applications.

Unlike the rest of our products, the Web Calculator is NOT registered as a Medical Device.

3. Getting Started with the EP Predict API

Access the API interface and Swagger documentation at https://epapi.endeavourpredict.org/.
Your key can be used straight away to request Risk scores.

4. Using the SDKs (Software Development Kit)

The portal provides downloadable components (.NET DLLs or Dynamic Link Libraries) that mirror the API’s functionality.

Key advantages
  • No network dependency
  • Faster execution
  • Full control over deployment
Typical workflow
  1. Download the library from EP Engines → SDKs.
  2. Add the DLLs to your project (full step-by-step guide, with code examples on our Wiki)
  3. Instantiate the PredictionService class using your key
  4. Pass in the required patient data, using an EP Input Model
  5. Receive a strongly typed result object called the Prediction Model

5. Processing Comma Separated Value (CSV) files with the Online Batch Processor

The Online Batch Processor is designed for high‑volume, non‑interactive workloads.

Capabilities
  • Generate Risk Scores for thousands of records
  • Export results for downstream analysis
Typical use cases
  • Population‑level risk stratification
  • Academic research
  • Service planning and forecasting

SDK Accounts also have access to an offline Batch Processor.

6. Web Calculators for Validation and Testing

The Web Calculator does not require a key, and developers can use it to:

  • Validate expected outputs
  • Compare API/library results
  • Understand model behaviour
  • Demonstrate functionality to stakeholders

This web calculator uses the same libraries found in the SDK, however it is NOT registered as a Medical Device.

7. Compliance, Safety, and Release Notes

Health‑risk models require strict governance. The portal provides:

Next Steps for Developers

  1. Explore the Web Calculators to understand model behaviour
  2. Review documentation and request an account and key
  3. Choose an integration path:
    • API for cloud or cross‑platform apps
    • SDKs for on‑premise systems
    • Batch Processor for large datasets
  4. Implement and test using sample payloads or test datasets
  5. Monitor release notes for model updates
  6. Deploy to production following your organisation’s clinical safety processes
©  Endeavour Predict