Innovation by ADOC Solutions
ADOC Solutions brings a specific vision to innovation. By definition, to innovate is to introduce something new. However, this notion of novelty is sometimes not totally foreign, it is sometimes a question of adapting processes to bring novelty.
Innovation to optimize processes.
Depending on the activities, it is sometimes necessary to optimize its processes to:
-
Create a different customer experience,
-
Reduce expenses,
-
Optimize your production times,
-
Improve the quality of productions,
-
Make operations more reliable,
-
Anticipate requests,
-
Build customer loyalty,
-
Develop new needs,
-
...
Artificial Intelligence - Analytics Technologies
Understanding an input flow is an essential operation to value the following processes.
What information do I receive:
-
Source,
-
Form,
-
Typology,
-
Expected,
-
Due date,
-
Recipients,
-
Opportunity & Risk,
-
...
How I understand this information:
-
By its form,
-
By its content,
-
...
How I make this information reliable:
-
By its relations to other documents,
-
Through consistency of content,
-
...
How I allow the engine to progress:
-
Manual Learning - Human Help,
-
Self learning,
Our Artificial Intelligence technologies are scalable thanks to the prior understanding of upstream information. Designing a self-learning architecture requires a complete understanding of the environment in order to be able to perfect the keys to learning. Developing a more or less complex neural network takes various shapes and times. They depend on the complexity of the content and the rules to make interpretations more reliable.
Some examples of Artificial Intelligence for the recognition and interpretation of content:
Automatic analysis ofsupplier invoiceswith extraction of line items. Understand an invoice type document (with the wide variety of forms for suppliers) in order to extract information from it, to bring it closer to other accounting content: orders, delivery notes, etc.
Particularity of the documents:
-
various forms according to the suppliers,
-
varied content depending on the activity (number of pages, several VAT rates, several billing centers, different languages, different currencies...)
Source of documents: email, paper, web portal...
If the first solutions deployed required a lot of modeling (by types of suppliers, etc.); today our tools allow self-learning without prior invoice modeling.
Automatic analysis ofRIS - Individual Status Reports.The statement of individual situation or RIS is an essential document to know your pension rights. It allows you to know the quarters acquired for the calculation of pensions.
Particularity of the documents:
-
varied forms according to the periods,
-
varied content depending on the activities,
Source of documents: email, paper, web portal...
​
​
Automatic recognition ofcellscancerous.ADOC Solutions participated for many months in the creation of an Artificial Intelligence (neural network capable of learning independently) for the recognition of cancer cells.
Different stages have been identified in the development of cancer: initiation, promotion and progression. Initially, there is a major lesion at the level of the DNA of a cell; this results in a transformation of this cell. In a second step, the transformed cell develops and proliferates by forming a group of identical transformed cells. Finally, in a third step, the cell acquires the characteristics of a cancerous cell: it multiplies in an anarchic way, partly losing its differentiated character (its identity linked to the tissue to which it belonged). The evolution is first locally, then can spread via the blood and the lymph to other places in the body where the metastases are formed. (source National Cancer Institute).
​
Automatic recognition ofPlants, Fruits, Vegetables.ADOC Solutions has developed an Artificial Intelligence (neural network capable of learning independently) for the recognition of plants, flowers, trees, fruits, vegetables... a SHAZAM for plants, fruits, vegetables... ADOC Solutions has developed a smartphone application to meet these expectations. Artificial intelligence, for this application, is based on a large knowledge base of all varieties that can be found. The evolution of the application (learning) resides in the construction of the engines, namely to interpret the variants of the objects.