Whether they cover a city or a whole country, our customers expect our products to be able to manage all their data with optimal performance.FIND OUT MORE with VEOLIA
On the ground: cables, equipment, connections. Our engineers see it as a graph, a mathematical object on which to apply powerful algorithms.FIND OUT MORE with the City of Paris
The Internet of Things is now a reality: lamp-posts, pedestrian crossings, and parking spaces can now communicate! Our products are there to listen – and make it possible to sift through the raw data to identify those objects that require particular attention.
From the field to the product
Our business experts identify the current and future needs of our customers. Our software publishing business consists of providing the best response to these needs as quickly as possible.
To achieve this, our development team uses a methodology called SCRUM: it is a so-called agile methodology, which ensures that we respond as quickly as possible to the requests of our customers.
In this method, every three weeks our product manager and the development team prioritise what needs to be covered and decide what the team will achieve in the next round. This allows us to adapt very quickly to our customers’ environment.
Also, every three weeks the development team delivers a version of the product in which the latest features are embedded: this allows us to check that what has been achieved is in line with the initial requirement.
Our development team consists of around thirties developers with varied profiles: from the student undertaking vocational training to the architect with twenty years of experience, they all bring their passion for technology, their sense of ergonomics and their business knowledge to our products.
The functional richness of our products is reflected in number of lines of code:
Delivery and training
Once all the tests have been passed, the product finally arrives on our shelves: the customer service teams then pick up the baton in order to disseminate the new version to our customers, train them and support them.
Because no robot can replace the expertise of our teams, with every release of a version the QA team screens each of our products on each of the supported platforms.
In total over 500 test scenarios must be confirmed, in addition to 700 hours of tests.
The quality of our products is our primary concern: it is the condition which is indispensable for the satisfaction of our customers.
No developer can keep an eye on 700,000 lines of code: our robots take care of this for us. Each function is read to ensure that it complies with all the rules and best practices. This quality measurement analysis enables us to detect or avoid certain bugs.
In parallel, the robots install the version of the modified product the day before and pass it to the test bench: more than 4,000 tests are carried out every evening, i.e. over four hours of tests where the functionalities are checked.
Our next challenges
- Graph modeling;
- Predictive : more from describing to predicting.
Our applications already know how to centralise all the data which is uploaded automatically or manually, whether they are reports of interventions entered by technicians, measurements taken by sensors or equipment self-diagnostics. Better still, our applications allow to use this information to make the work of our customers more efficient: if a particular router indicates that it is out of order, which connections need to be modified to restore the subscribers’ connection? The power consumed by a particular electrical cabinet has just failed, would a lamp it supplies not have burnt out? Why not go further? How can one predict breakdowns from the uploaded data? How can we detect recurring patterns in the use of public facilities, in order to improve their availability?
Our technical partner
As a global leader in databases, Oracle has rolled out its products in more than 145 countries around the world and has 400,000 customers, including the 100 largest US companies. All GiSmartware products are compatible with Oracle databases.