Patented methodology gives data scientists more power to make rapid iterations without having to involve the application team
Datatron has received a new U.S. patent for its methodology for modeling machine learning and analytics, the company announced today. This expands Datatron’s patent portfolio for machine learning (ML) methods and technologies in a production environment, including data extraction, training of the model and predicting of the module. To view the patent, visit https://tinyurl.com/ejnh4vph.
Click to Tweet: @datatron announces latest patent for machine learning models https://tinyurl.com/ejnh4vph #MLOps #ArtificialIntelligence
ML introduces operational complexities and risks that need careful attention, and many projects fail due to siloed roles and disjointed processes. Datatron’s newly patented technology builds on the company’s track record of enabling businesses to easily, accurately and rapidly operationalize and monitor AI and ML models in production. It includes the introduction of a “challenger gateway module” that stabilizes the model’s endpoint and provides a bridge between the application team and the data science team.
Data scientists change models and experiment with new models, but in most production environments, they’re siloed from the application team. By bridging the divide between these two teams, Datatron’s methodology gives the data science team more flexibility in terms of controlling model changes and experimenting faster in production. Datatron customers are already using the “challenger gateway module” to achieve improved results.
Harish Doddi, CEO, Datatron, said: “When it comes to MLOps, it’s the data science team that controls the models, but the application team that controls the application release. It’s a constant struggle requiring constant coordination. With our patented methodology, we are giving more power to data scientists and reducing the frustration of the application team. That sets up the organization to truly realize the benefits of their machine learning projects.”
For more such updates and perspectives around Digital Innovation, IoT, Data Infrastructure, AI & Cybersecurity, go to AI-Techpark.com.
