For the latest on new releases or to become a Design Partner to share your input, sign up below.
Execute a Couple of Commands:
build and apply
torque deployment build local
torque deployment apply local
torque deployment build prod
torque deployment apply prod
Create Deployment Objects by Selecting a List of Deployment Provider Implementations
torque deployment create local torque.docker_compose.V1Provider
torque deployment create prod torque.do.V1Provider torque.k8s.V1Provider
Create a DAG Representing Your Cloud Architecture
Create Torque Workspace
torque init my_workspace
To Deploy to DigitalOcean, Kubernetes, and Docker Compose
An Example of Torque at Work
Kubernetes and Terraform, coupled with all services that cloud providers offer these days, have provided incredibly powerful tools for running cloud systems. However, the level of complexity they introduce can be daunting for many teams, often creating technical debt accumulation, slowdowns, and roadmap disruptions for refactorings or rearchitecting.
Torque allows engineers to design and easily evolve their systems without being constrained by fragile manual configurations of development environments and cloud deployments.
Being removed from static configuration files gives software engineers the power to deliver new features fast without breaking things or accumulating technical debt.
Why Torque Framework?
Tooling for (Data|ML|Dev)Ops
We specialize in delivering
built on top of an open-source framework powered by libraries
By installing different libraries you can:
quickly change datasets
simply modify data pipelines
change connected data sources
keep track of data-to-model relations and executions
deploy models to staging and production environments
rely on your system dependency graph as a single source of truth to plan and execute changes
Because Torque is technology agnostic, and can be adopted one component at a time, getting started does not require a significant change to your current system design, preferred tools, or development practices.
Why Use Torque Automation Framework?
Countless tools and services offer incredibly powerful features that serve various needs through the software development pipeline.
However, the level of complexity they introduce can be daunting for many teams, often creating slowdowns and delivery disruptions.
Unlike static scripts, components are effectively self-aware and self-configurable, meaning your system can be built and deployed to any environment with any required functionality that your installed code libraries support.
You can even run any of your favorite tools, or deploy to your favorite platforms, through libraries, so you gain their functional benefits without manually managing their code.
What is Torque Automation Framework?
Built specifically to simplify software system operations, Torque is a framework that, when installed in your codebase, speeds up model deployment and learning cycles.
Instead of making changes to every stage of operations by editing static scripts and configuration files, you use simple commands that modify your system dependency graph.
Through installed code libraries that describe how the components of your system behave throughout all stages, Torque turns each new change into concrete automations, from the local development environment, over CI/CD, all the way to deployment configurations.
Whenever you need to, install different open-source libraries, fine tune each implementation to suit your own needs, or write your own from scratch.
Torque Framework allows you to iterate faster by:
operating on a higher design level, using simple commands to add or change components that also know how to interconnect
installing open-source libraries that implement all stages of each component from development to build pipeline to deployment
reducing repetitive manual tasks by customizing libraries, to make changes dynamically for retraining and re-deployment