Work in Progress (Demoware -- series of posts & talks coming to show how I built this)

Introduction

A solution that uses ML.Net to classify spending patterns. Utlising Azure features to be self updating (aka online machine learning)

For more info

Charge Id – scratching the tech itch [ part 1 ]

Charge Id – lean canvas [ part 2 ]

Charge Id – solution overview [ part 3 ]

Charge Id – analysing the data [ part 4 ]

Charge Id – the prediction model [ part 5 ]

Charge Id – deploying a ML.Net Model to Azure [ part 6 ]

Code

https://github.com/chrismckelt/vita

Technology

• ML.Net classifier for NLP description [done]
• ML.Net regression for unclassified transactions [to do]
• Azure Service Bus & Queues to update the prediction model [to do]
• Azure app service with a Swagger REST API to run the predictions [done]
• Azure FUNCTION app service with a Swagger REST API to run the predictions [blocked]
• Angular 6 with SSR [doing]
• Application Insights [to do]
• local docker dev with CosmosDB [to do]
• CosmosDB backing onto Azure Search for quick lookup of data (Australian companies, suburbs and Google place information) [doing]

Project board

https://github.com/chrismckelt/vita/projects/1

License

MIT.