JAVLN is a fast growing NZ insurtech, solving broker pain points with smart software. International market entry is a strategic objective, some of which need new intelligent software capability to be unlocked. JAVLN was introduced to Arcanum through AWS.
Led by CEO Dale Smith, it took only one call for JAVLN to validate the need for machine learning to support their journey into new markets. JAVLN needs to focus all available developer resources on the core application; working with a partner to build and support a new machine learning capability fit the need for speed and minimal resource distraction.
Extracting 10 pieces of key information from insurance documents is the core capability JAVLN needed to secure new customers.
- Our data science team provided a plan to build, deploy and run the models; including a calculator for the AWS running costs.
- Arcanum built a solution to; Run text extraction (OCR), Pre-process raw text, An NLP model to produce a final output
- The Arcanum Accelerate platform is used to create deployment code that will run in the JAVLN AWS environment.
- To support faster integration, Arcanum built a simple prototype web app to test inputs and outputs.
- Once integrated with the core app, detailed MLOps performance monitoring will be available in the Accelerate platform dashboard.