Elastic Infrastructure for AI
Available on any public cloud or on-premise.
Instantly deploy machine and deep learning models as production-ready, self-healing, auto-scaling microservices. Mix and match frameworks without losing momentum.
Eliminate duplicate work by housing all your microservices and AI models in one place, from across business lines and technology stacks. Thousands of services, one dashboard.
Multiply the contributions of your best talent by making their work available to every corner of your company. Build up and grow your algorithmic moat and data strategy.
Use Git and write functions in Java, Scala, Python, Ruby, NodeJS, Rust, or R. Codex is stack-agnostic and compatible with your current workflows and IDE. The REST API nature eliminates any interoperability challenges.
Combine, pipe, and remix functions to create new solutions. Codex allows you to connect functions from different technology stacks. Individuals and teams will build on each others’ work instead of re-inventing the wheel.
Categorize, tag, and describe functions in a searchable format. Individual contributors are able to find, evaluate, and integrate the work of their peers.
Publish functions as versioned API endpoints. Codex makes each version of a single function available as a unique REST API endpoint, concurrently, eliminating backward-compatibility challenges and breaking changes.
Cherry-pick the latest algorithms from the public marketplace that are written by world-class researchers. Instantly expand the tool chain of your internal developers and scientists.
Configure Codex to automatically scale up or down in response to fluctuating throughput needs. Substantially reduce data center costs by avoiding over-provisioned clusters.
Horizontally deploy and scale deep learning models over GPU clusters. Effortlessly transform complicated models to universal REST APIs.
Monitor cluster-wide health metrics and user-specific audit trails. Codex can be configured to meet the most stringent data security policy.
Codex is the technology behind the largest algorithm marketplace in the world and offered with 24x7 hands-on engineering support.