Key results
The company
KOHO
koho.caDigital banking platform for spending, saving, and credit building in Canada.
Result highlights
- 50% reduction in security resolution time
- 3x faster AML report validation
The story
A leading Canadian fintech serving over 2 million users with spending, savings, and credit-building solutions built entirely on cloud infrastructure.
Strict financial data privacy standards made it difficult to scale generative AI across operations without compromising compliance. Isolated tools lacked the centralized governance needed for sensitive workflows like anti-money laundering investigations and security event resolution.
The organization developed Kortex AI, a centralized internal platform that provides secure access to foundation models via Amazon Bedrock while enforcing strict governance controls like model approval lists. Amazon Aurora serves as a vector database to power retrieval-augmented generation (RAG) capabilities for context-heavy tasks like lease validation and regulatory reporting. The solution integrates directly into employee workflows through a web interface for general analysis and IDE plugins for developer coding assistance.
Scope & timeline
- Dark mode feature built in 2 days
- 120-140% higher pull request throughput
- 66% increase in deployment frequency
Quotes
“Our goal was to provide KOHO employees with better, easier, and more secure access to large language models.”