← Back to case studies
Ubix.io
Enhancing Ubix.io's Analytics Platform with a Custom Domain-Specific Language
Analysis time
baseline
→
−90%
Who can explore data
data scientists
→
domain experts
Ubix Engine is a big data science platform that enables data science automation and advanced analytics applications. Ubix provides data, metadata, and analytical services for both historical and real-time streaming data under a unified lambda architecture.
What changed
Before
Request
Analyst asks
Bottleneck
Data scientist
writes the code
Result
Days later
After
Interface
Custom DSL
no programming needed
Engine
Spark
Cassandra
Elastic Stack
Delivery
Self-service
−90% time to insight
A domain-specific language put exploration in the hands of analysts, so the data-science team stopped being the queue.
What we built
Ubix, an analytics company, wanted to enable their domain experts (non-technical users) to efficiently explore data and build predictive models without writing complex code. Data scientists also required a more advanced framework for creating sophisticated models quickly. We built a custom Domain-Specific Language (DSL) that simplified data exploration, feature engineering, and model building — intuitive for non-programmers but flexible enough for technical users.
Key Features
Built a DSL system that enables non-technical users to explore data, perform feature engineering, and build predictive models with simple, human-readable commands.
Enabled data scientists to use advanced modeling techniques and integrate with existing tools for deep analysis.
Provided a web-based interface for running DSL commands, generating visualizations, and creating reports.
Ensured scalability and support for large datasets in a distributed environment.
Results
Analysts who previously relied on data scientists could now explore datasets and build models independently using simple DSL commands, reducing time spent on data analysis by 90%.
Domain experts were able to perform advanced data exploration and modeling tasks, leading to better insights and decision-making without requiring programming skills.
Implemented a fully automated CI/CD pipeline to support continuous deployment.
Configured monitoring, alerting, and automated scaling to enhance operational efficiency.
Designed product infrastructure architecture on AWS and implemented automatic deployments using IaC.
Stack
Engagement: 2 years · Real-time & Industrial Data Platforms, Platform & Cloud Engineering