Usermind required a proven, complete, web-scale and enterprise-grade solution as means to accelerate its own data science offering, intellectual property & team. The full enterprise-grade data science platform provided by John Snow Labs includes a full deployment to development and production environments, security hardening, integration into the Usermind platform, training and support. The platform is fully white-label, Usermind has its full source code and the license to modify and extend it at will.
The delivery process included both the software product and a turnkey service that ensured Usermind was in production successfully and that its team was familiar and confident with the source code and operational aspects.
“John Snow Labs assembled a multi-continent team of big data experts that allowed Usermind, within three months, faster than we thought possible, to build several significant components that are providing real value to our customers.” – Chris Jones, Principal Engineer, Usermind
“Just gave demo of everything to our first customer. It was a home run. Thanks to all.” – Michel Feaster, Usermind CEO & Founder
“John Snow Labs brought expertise in a variety of technologies including ElasticSearch, Spark, JupyterHub, OpenScoring and Machine Learning, significantly accelerating Usermind’s development and expertise.” – Chris Jones, Principal Engineer, Usermind
Get your Data Science Platform from concept to production in weeks
A good data science platform makes it easier to find and understand past analysis, so that data scientists don’t need to start from scratch when asking new questions. It scales to explore data on multiple machines, without major DevOps or infrastructure setup. Most importantly, it includes the use of new analytics packages and tools, safely – without breaking past work or disrupting environments for colleagues. These challenges have all been solved by John Snow Labs with its data science platform, which can get your enterprise ready to make more revenue in just a few weeks.
Data Science Platform Components
Data Discovery and Visualization components are based on ElasticSearch and Kibana. It allows you to get the best user experience while exploring data. This platform provides functionality to interactively browse data, create different charts and other visualizations based on stored data and assemble custom dashboards.
Collaborative Data Science is based on JupyterHub. This is a multi-user server for notebooks that provides ability to create and share documents, analyses and models between scientists.
Model Scoring is based on Openscoring, and extends it to provide security, scaling, fault tolerance, monitoring and logging. The model server provides a fast, enterprise-grade solution that allows you to serve multiple models simultaneously.
Other components of the platform include identity management, monitoring, backups, cluster processing, data curation, scheduled workflows & dataflows.
John Snow Labs’ Data Science Platform’s Benefits:
- Enterprise-grade solution – Production ready out of the box with strong security, monitoring, failover, backups, single sign-on, auto-scaling and self-healing.
- Cutting Edge Technology – No compromise inclusion of the latest certified open source libraries in deep learning, NLP, reinforcement learning, distributed machine learning and others.
- Turnkey Deployment – We get you to production quickly & reliably. We do the first deployment, configure and verify all components, and train your team to take it on.
- Self-Sufficiency – Knowledge sharing sessions to train your team on the source code, architecture and operations
- Intellectual Property – The platform is white label. You get the full source code, the right to modify and extend it at will, and whatever you add is yours
- Petabyte Scale – Every component is designed and verified to scale out horizontally on commodity hardware