Deploying EnterpriseWeb
Lightweight, scalable and resilient.
EnterpriseWeb implements a lightweight, horizontally architected, serverless architecture designed for high performance, scalability and resilience.
The cloud-native harness deploys as a cluster of pods with a small footprint (50 MB). It can run on-premise, in the cloud, at the edge, or even on a laptop. Serverless processing scales out naturally over allocated resources.
EnterpriseWeb is also available as a Snowflake native app that deploys to Snowpark containers within a customer’s tenant under Snowflake’s world-class Security Hub and Horizon data governance. EnterpriseWeb’s Snowflake native app also comes with rich ready-to-use integrations to Snowflake data, analytic and AI services so customers can quickly operationalize data in their AI Data Cloud.
EnterpriseWeb is API-first. It also offers a graphic user interface for its design and management environments. In addition, EnterpriseWeb offers a REST/CLI console for NLP interactions with agents and language models.
As a cloud-native platform with a serverless runtime, the Semantic AI Harness naturally decouples compute from storage, late-binding everything for runtime contextualization.
EnterpriseWeb doesn’t use a Graph Database or Triple Store; it can be implemented over virtually any database, eliminating the need for specialized databases and database administrators. The platform is best suited for column family databases, which are designed for high-performance read/writes. EnterpriseWeb is generally bundled with Cassandra along with SOLR for indexing.
EnterpriseWeb’s runtime handles all higher-level transaction controls and state management, as well as logging and tracing. It abstracts the persistence tier for developers, eliminating database administration concerns.
Column family databases are one of the most basic, efficient and compressed forms of storage so they scale up naturally. Since EnterpriseWeb abstracts database management and Cassandra supports distributed architectures, the platform can also scale-out without sharding overhead or complexity.