CoherentPaaS - Coherent and Rich PaaS with a Common Programming Model

What user need or pain point is your project addressing?

I am the head of the IT department of a large enterprise. Just like many other companies, we are using a combination of data management solutions, such as SQL databases, NoSQL databases such as document data stores, graph databases, key-value data stores and complex event processing. However, we are encountering some problems as some business actions require us to update multiple data stores. In addition, whenever there is a failure our federation of data stores loses its consistency in dealing with multiple problems with our customers.

What is more, we now have to perform operations across our heterogeneous data stores. This is a big problem for us as when we have to join data, our application developers have to develop the code. Most of the time they are not able to do this. This implies that we also have to move the data from the data stores to our data warehouse in order to be able to perform these queries. This is also painful because the data warehouse is a SQL OLAP database and does not support the graph and document data models for either of the corresponding query languages.

Project's major results: 

A polyglot persistence platform that provides queries and transactional updates across SQL and NoSQL data stores.

Target stakeholders: 

Small & medium enterprises, Large companies.

Find out more

Case Studies

Going Global - CoherentPaaS Position Paper for Cloudscape Brazil 2015, Ultra-Scalable Cloud Database

Project Start: 
Project End: 

How will your solution/service benefit the end-user?

CoherentPaaS addresses my two pains points. Now we have transactional semantics to update across data stores getting all-or-nothing semantics for business transactions which guarantees our data consistency in the advent of failures. This means no more problems with customers due to data lost during failures!!!!

CoherentPaaS has also successfully solved our pains with making queries across data stores. Now, we can write subqueries in the native query languages, exploiting the full power and efficiency of the underlying data stores, including their data models and query languages/APIs. When we have queries across data stores we write them declaratively in SQL. This SQL observes the resulting sets of the subqueries as temporary SQL tables. With CoherentPaaS, now we do not have to waste any extra time and effort on the heterogeneity of our data stores and we can fully exploit the data.

CoherentPaaS has helped us to solve two major pains points we encounter due to the introduction of multiple data store technologies in the company and has reduced the huge overhead in terms of personnel effort that we previously had to perform to access the data scattered across data stores.

Potential exploitation strategy: 

The exploitation strategy has lied in the creation of a spinoff of the project, LeanXcale, incorporated in mid March 2015.

Vertical Market: 

Smart cities, Energy, Finance & insurance.