{"id":"47380b25-1852-453e-b6da-26c34cd4de96","createdAt":"2024-07-16T23:00:00.000Z","updatedAt":"2024-08-01T18:28:04.916Z","isDeleted":false,"version":271,"date":"2024-07-16T23:00:00.000Z","published":true,"title":"Serverless Model Serving: Optimize AI deployment","notes":{"type":"doc","content":[{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"The ","type":"text"},{"text":"Serverless Model Serving","type":"text","marks":[{"type":"bold"}]},{"text":" feature on ⚗️ ","type":"text"},{"text":"Instill Model","type":"text","marks":[{"type":"link","attrs":{"rel":"noopener noreferrer nofollow","href":"https://www.instill.tech/docs/model/introduction","class":null,"target":"_blank"}}]},{"text":" offers a more cost-efficient solution for AI deployment by eliminating the need to manage the underlying infrastructure. ⚗️ ","type":"text"},{"text":"Instill Model","type":"text","marks":[{"type":"bold"}]},{"text":" handles the compute resource provisioning, auto-scaling and availability, allowing you, the developers to focus on model development and deployment 🎉.","type":"text"}]},{"type":"paragraph","attrs":{"textAlign":"left"}},{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"Key Benefits:","type":"text","marks":[{"type":"bold"}]}]},{"type":"bulletList","content":[{"type":"listItem","content":[{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"Scalability:","type":"text","marks":[{"type":"bold"}]},{"text":" Automated resource allocation scales to handle varying workloads efficiently.","type":"text"}]}]},{"type":"listItem","content":[{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"Cost-Efficiency:","type":"text","marks":[{"type":"bold"}]},{"text":" Compute resources are provisioned on-demand and scaled down to zero when idling.","type":"text"}]}]},{"type":"listItem","content":[{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"Simplified Management:","type":"text","marks":[{"type":"bold"}]},{"text":" No need for server management, free up your teams to focus on improving models and applications.","type":"text"}]}]},{"type":"listItem","content":[{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"Quicker Deployment:","type":"text","marks":[{"type":"bold"}]},{"text":" Seamless integration with 🔮 ","type":"text"},{"text":"Instill Core","type":"text","marks":[{"type":"link","attrs":{"rel":"noopener noreferrer nofollow","href":"https://github.com/instill-ai/instill-core","class":null,"target":"_blank"}}]},{"text":" for faster full-stack AI development with data pipeline and knowledge bases.","type":"text"}]}]}]},{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"We've also made significant performance enhancements:","type":"text"}]},{"type":"bulletList","content":[{"type":"listItem","content":[{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"Faster image push times","type":"text"}]}]},{"type":"listItem","content":[{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"Reduced model cold-start times","type":"text"}]}]}]},{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"We’re continuously enhancing ⚗️ ","type":"text"},{"text":"Instill Model ","type":"text","marks":[{"type":"bold"}]},{"text":"and will soon introduce ","type":"text"},{"text":"Dedicated Model Serving","type":"text","marks":[{"type":"bold"}]},{"text":" for production-purpose model serving service.","type":"text"}]},{"type":"paragraph","attrs":{"textAlign":"left"}},{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"☁️ ","type":"text","marks":[{"type":"bold"}]},{"text":"Instill Cloud","type":"text","marks":[{"type":"link","attrs":{"rel":"noopener noreferrer nofollow","href":"https://instill.tech/hub/featured/pipelines","class":null,"target":"_blank"}},{"type":"bold"}]},{"text":" users can host public models for free and we offer monthly free ","type":"text"},{"text":"Instill Credits","type":"text","marks":[{"type":"link","attrs":{"rel":"noopener noreferrer nofollow","href":"https://www.instill.tech/docs/cloud/credit","class":null,"target":"_blank"}}]},{"text":" for ","type":"text"},{"text":"☁️ Instill Cloud","type":"text","marks":[{"type":"bold"}]},{"text":" users to run any public models hosted on ","type":"text"},{"text":"☁️ Instill Cloud","type":"text","marks":[{"type":"bold"}]},{"text":".","type":"text"}]},{"type":"paragraph","attrs":{"textAlign":"left"}},{"type":"paragraph","attrs":{"textAlign":"left"},"content":[{"text":"More at: ","type":"text","marks":[{"type":"italic"}]},{"text":"Instill AI","type":"text","marks":[{"type":"link","attrs":{"rel":"noopener noreferrer nofollow","href":"https://bit.ly/3A8ihGc","class":null,"target":"_blank"}},{"type":"italic"}]}]}]},"archived":false,"imageUrl":"https://uploads.productlane.com/c3193d03c7efe0f5f66619bc1af5cb00.png","projectId":null,"changelogStatus":null,"workspaceId":"52f06d0d-2381-411e-a8c5-7b375e3a0114","html":"

The Serverless Model Serving feature on ⚗️ Instill Model offers a more cost-efficient solution for AI deployment by eliminating the need to manage the underlying infrastructure. ⚗️ Instill Model handles the compute resource provisioning, auto-scaling and availability, allowing you, the developers to focus on model development and deployment 🎉.

Key Benefits:

Scalability: Automated resource allocation scales to handle varying workloads efficiently.
Cost-Efficiency: Compute resources are provisioned on-demand and scaled down to zero when idling.
Simplified Management: No need for server management, free up your teams to focus on improving models and applications.
Quicker Deployment: Seamless integration with 🔮 Instill Core for faster full-stack AI development with data pipeline and knowledge bases.

We've also made significant performance enhancements:

Faster image push times
Reduced model cold-start times

We’re continuously enhancing ⚗️ Instill Model and will soon introduce Dedicated Model Serving for production-purpose model serving service.

☁️ Instill Cloud users can host public models for free and we offer monthly free Instill Credits for ☁️ Instill Cloud users to run any public models hosted on ☁️ Instill Cloud.

More at: Instill AI

Instill AI changelog

Serverless Model Serving: Optimize AI deployment