Manta
How Coding Bear designed and developed Open Manta Designer, a tool for modeling and visualizing data flows
Intuitive visual modeling
Users can quickly and easily create data flows using drag & drop, enrich existing metadata with additional information, visualize complex structures, and reflect all changes in the central shared repository.
Challenge
Manta needed to enable less technical users to manually create and edit data flows, and to enrich the central repository — built through regular automated scanning of source systems and databases — with this metadata.
Data lineage at large organizations spans a wide range of technologies and data sources, so the tool had to be flexible, easily extensible, and ready for new types of systems without requiring code changes. Open Manta Designer was meant to enable intuitive lineage modeling, metadata enrichment, and quick integration of custom elements in areas where Manta's automatic scanners cannot provide full coverage.
Working on something similar? Reach out and we'll connect you with specific references.
Free consultation→What CodingBear suggested
Coding Bear designed and developed the Open Manta Designer tool end-to-end: from requirements analysis through architecture and UX design to implementation and testing. The tool is built as a configurable system, making it easy to add new types of data sources, extend models, and adjust application behavior — all without touching the code.

After a thorough analysis of available visualization libraries, we chose Konva.js, which provided the flexibility and performance needed to work with complex data structures. This allowed us to implement multiple drawing canvases with data synchronization between them, maintain full control over the appearance and behavior of visualized elements and curved connectors, and add undo/redo functionality along with basic model validation. The rendering algorithms with real-time calculations handled manipulation of hundreds of elements and thousands of connectors within a single canvas — without any significant performance impact on the application running in standard browsers.

The frontend and visualization layer were written in React, and the entire solution was cloud-ready from the start.
We delivered the solution in full scope, and thanks to close communication with the Manta team, we were able to continuously align priorities and individual steps. The project became part of the platform's strategic development — which subsequently strengthened its position even after joining IBM.
%25201.webp)