Key features and benefits

The unique combination of Sesam's key features leads to simplified management of master data in data driven environments. The service includes most of the core features needed to build a data-fabric based platform.

Delta stream processing

Delta processing means that only data that has changed is ever processed by Sesam. The delta stream processing pipes collect raw data, connect data objects across sources, and store data into a combined semantic data storage. The same delta stream processing pipes transform and share the data with any target system, using each systems’ native schema. Thus, all data processing in Sesam is unified in a low-code, high-throughput, low-latency service.

Benefits of delta stream processing

Sesam processes only deltas, so that no systems connected by Sesam need to reprocess existing data. This reduces system processing time and update delays by at least 95% compared to bulk ETL jobs, and completely removes unnecessary data updates.

Graph-based dependency tracking

Graph-based dependency tracking ensures that every composite object is always updated as a delta stream throughout the Sesam service, no matter the number of complex transformations, including merging from any number of source objects. This dependency tracking is made possible by the fact that there is no code in Sesam, only metadata configuration describing the transformations in pipes.

Benefits of graph-based dependency tracking

Sesam supports graph-based dependency tracking, so that even complex objects from multiple sources can be processed as delta streams. This ensures that complete reprocessing is never required, no matter the number of sources, the complexity of the transformations, or the number of target systems and data schemas involved.

Sesam uses metadata configuration, so that all data flows uniformly, without the need for custom code, giving complete transparency and heavily reduced operating cost.

Semantic data storage with dynamic multi-schema

In Sesam the output of any delta stream processing pipe is stored in semantic datasets. A semantic dataset is an immutable object store, capable of containing the complete version of any object. Every dataset has a completely dynamic schema, continuously updated to fit the data stored in it. Semantic technology ensures that each property´s origin is preserved and allows a single object to contain multiple schemas, assembled into one global object. Adding golden properties, and continuously evolving them, can be done whenever needed, as these properties are added to the existing global object without affecting the original properties.

Benefits of semantic data storage with dynamic multi-schema

Sesam stores all data in semantic data storage, so that data never loses its origin and context, and never gets corrupted or lost in translation. Every representation of an object, regardless of source, is stored as a single entity, without conflict or consistency corruption.

Sesam connects every representation of any single object, and stores it as a multi-schema global object, so that any source system only needs to share its data once, and in its native schema. On average this reduces the number of integrations by 90%, in addition to simplifying each integration by never having to modify any source systems or do any data modeling before adding source data.

Sesam supports continuously evolving global properties in the multi-schema datastore, so that canonical models can emerge over time, and any standardized data model can be implemented independently and be stored in a single object. When all source properties and global properties are available as a single composite object, the cost and time of acquiring data and transforming the data into a target system in its native schema is heavily reduced.