Multiple Data Scientist with leading Airport software provider developing models for different languages and framework.
Platform required where the Models can be managed, re-trained, scaled for data, deploy in production with different environment.
Using open source MLOps tools, built a AI workbench where models can be registered, re-trained and deployed with a seamless UI feature.
Provides ability to scale on large data & models can run in real time. Logging feature and access control been provided. Complete API led development.
- Improvement in collaboration between data scientist resulted in cross pollination increasing 40% reusability
- Model management reduced from days to hours
- Scalability, better operations handling and quick deployment across multiple platform improved GTM