Methodology and architecture
Using an agile methodology, the system will be implemented as a
set of microservices, integrated into managed containers. Everything can
therefore be perceived as a series of black box components, connected by
well-defined input/output parameters.
For the MVP/prototype, some black boxes will contain calls to
generative AI applications. However, these components will gradually be
replaced by our own AI (e.g., using neural networks, machine learning, etc.).
Preparation of system documentation.
To speed up development by the SOLSER team, the original system
architect (Pierre) will soon provide complete documentation of the analysis and
design.
This will be subject to review by SOLSER's Scrum Master (Abraham)
and lead architect (Irving) to ensure that it complies with the company's
practices.
Since this is an agile project, the documentation will evolve as
development progresses. However, the development team should soon have all the
information they need to proceed.
The SOLSER development team
In addition to the Scrum Master and the company's chief architect, it will take a diverse development team to implement this product correctly.
Ideally, this would include people with knowledge of:
- User interface development,
- Database programming,
- Big Data,
- Business logic,
- Generative AI applications and
- Deploying microservices using containers.
Some of the relevant skills include Angular/React, SQL/NoSQL,
C#/Java, and Docker/Kubernetes.