Start deploying your data science code effortlessly.
UbiOps is a deployment, serving and management backend for your data science and data processing code. It is there to help you create production-grade data-driven applications with ease.
An introduction to UbiOps
Deployment
You are in charge of your code. UbiOps takes care of the rest.
- Leverage the power of Docker and Kubernetes without the hassle
- Run your code as a scalable microservice with an API serving endpoint
- Make use of external libraries, files, artifacts, or other dependencies in your code
- Low latency synchronous requests & asynchronous batch requests.
- Out-of-the-box automatic scaling
- Intuitive model version management for deployment updates and rollbacks.
- Logging & Monitoring
- Available as SaaS or on-premises in your own environment
Connect your data & Construct data flows
Construct data flows with multiple processing steps. Connect to data sources and sinks.
Create applications by combining multiple deployments in a data flow. Connect your data transformations, predictions and pre/post-processing of data. Let UbiOps manage the heavy lifting.
Serve pipelines behind their own unique API endpoint.
Many data input & output options. Work with JSON, images, text files, plain text or many other data types.
Create applications with multiple processing steps using pipelines.
Connect with databases and storage systems to create managed end-to-end data flows.
Schedule workloads and pipelines with ease.
INTEgrate with your stack
A layer that fits right in with other tools and services in your stack.
- Make use of our Client Library and CLI to integrate with UbiOps
- Have all flexibility to use the Python libraries and frameworks you need
- Templates and examples for connecting to common data sources and integrating with cloud services and other tooling
- A layer between your data science code and IT infrastructure
risk-free operations
UbiOps helps you to remove the risk of setting up, securing and maintaining infrastructure.
- Separation of concerns between data science code and infrastructure
- Setting up & configuring production infrastructure yourself can be tricky. We did the work for you
- Leverage the power of Kubernetes and Docker without the hassle of maintaining them
- Fail-safe and self-healing deployments
- Monitoring & Logging functionality to keep track of what is happening
Governance & Control
Keep the overview on your applications and workloads from a single interface.
Secret and credential management. No need for hardcoding passwords in your code.
Keep track of requests and their status.
View logs from your code and from the UbiOps infrastructure.
Extensive permission management.
SSO authentication.
Get metrics on usage and performance.
Run anywhere
Make use of our SaaS, or run UbiOps in your own cloud environment or on-premises.
You are in good company
Many teams are already deploying smarter with UbiOps
“What I mostly appreciate are the fast response times and user-friendly UI. Have a new database, schema or storage platform? Just add a connector to the pipelines. Want to conduct live A/B tests? Just add another model to the existing pipeline, and run it without compromising the existing infrastructure. ”
Enrique Gutierrez, Data scientist at ASSET Rail
“Using UbiOps, we manage to reduce our costs by 85%. We don’t have to worry about the set up, nor the continuous maintenance of our models and pipelines, even when things get more complex”
Mart Folkerts, Head asset management at ASSET Rail
“We can quickly conduct pilots, and if successful, deploy it in production. We use UbiOps for different types of scripts: data cleaning/reshaping, several regression models and one neural network. ”
Mart Folkerts, Head asset management at ASSET Rail
“Examples provided there are easy to follow and very practical. With this documentation, it is possible to use UbiOps without getting any information up front.”
Stephanie Luik, Data Engineer at BAM
“UbiOps enables us to develop, deploy and operate any type of data science code, without having to worry about the IT infrastructure. Even when we continue growing in size and amount of data science applications.”
Pieter van der Mijle, Data Scientist at BAM Energy Systems
“Building such an infrastructure ourselves in our own Cloud would cost us considerably more in terms of money and time. With UbiOps we are able to fasten the time to market of our models with 50%”
Peter Visser, IM Enterprise Architect at BAM
“Ubiops helped us to deploy our product to a production environment in a scalable but cost-effective manner. They platform works like a charm and is a joy to work with for any data scientist ”
Joris Mollinga, Co-founder at DuckDuckGoose
“By collaborating with UbiOps we can efficiently serve our machine learning models on their platform. Their platform allows us to reach a larger audience in a very efficient way. ”
Joris Mollinga, Co-founder at DuckDuckGoose
“We have a critical infrastructure using sensitive data from smart meters.
Consequently, our standards in terms of reliability and security are very high. UbiOps meets and even exceeds these standards. ”
Ivo Fugers, AI Lead at Gradyent
“Using UbiOps, our team is able to orchestrate all processing steps and data flows seamlessly so we could start using our models while minimizing DevOps costs.”
Hervé Huisman, Founder & CEO at Gradyent
“With UbiOps we managed to significantly reduce the time and efforts needed for deploying and managing our AI models for detecting failures of our switch heating installations”
Cees-Jan Mas | Projectmanager Vernieuwing at ProRail
