After several iterations with our sponsor, we are very proud to announce the new topics for the 24-25 academic course. As usual, they are hot topics in technology and development, completely focused in the business core of the service operator. You will find the grants listing at the end of this page, but before this, please, take a look into this important information:
Our grants are designed to select the best talent from our University, taking the compromise of developing them both technically and also professionally in their soft-skills. We will help to achieve together the following objetives:
Open the corresponding folder to have a bried overview of the project and also the target team in Mas-Orange’s structure. The final M+O destination team could change depending on the company needs.
As you will probably need more information for all of them, you should consider reading this blog post which describes the Mas-Orange IT fundamentals. After reading it, if you have additional questions, don’t hesitate to ask them in the email: alvaro.paricio@uah.es indicating in the subject «More info for the MANEDS project XX-XX-XX».
Of course, the sooner you apply for the grant, the sooner we will consider your application. You can apply at this link: Apply to the Catedra MANEDS (MAS-ORANGE) project grants
It is also important to consider that:
- We are not looking for experts. You will learn with us and hopefully, become one.
- Your attitude is your best tool.
- You won’t work alone. We always work as a team, providing cross-support one to each other.
- You will have two tutors: one of the Chair teachers, and one professional tutor in the professional team of the company.
24-01-IA / Genetic Prompting over Generative AIs
M+O Target: AI Team
Technologies to learn: AI Tools, python, Matlab, Jupyter/Anaconda.
Target Profile: IA Engineer
Generative AI is a powerful trend that has raised a new competence: the prompt engineering, as the set of rules that enable creating the best possible question to the generative engines. Nevetheless, how could we asses that a prompt is optimal? How could we refine the prompt ina a parametric way?
It is possible to create parametric prompts that can be evolved using genetic algorithms that enable finding the optimal response. This project is focused on designing a new research on evolutive prompting using automated methods. These prompts will be used for multiple purposes such as coding, testing, reporting, finding best choices, assessing, etc.
You will:
24-02-BK / Advanced Microservices Architectures for the Multi-Service Back-Office
M+O Target: IT Architecture Team
Technologies to learn: Java, python, Swagger, Jupyter/Anaconda, Vertx.
Target profile: Service Architect
Investigate the advantages and disadvantages of using HTTP REST APIs vs. gRPC in a microservices architecture developed in Vertx. The goals is to identify and clarify the implementation patterns in the platform, to distinguish when to use each of the API technologies, considering the strengths and weaknesses of each one when implemented in Vertx.
Implement example cases for both.Analyze and put into practice the implementation of each case in a microservices monorepository.
You will:
24-03-BK / Serverless Architectures for the Multi-Service Back-Office
M+O Target: IT Architecture Team
Technologies to learn: Google-Cloud, AWS Lambda, Docker, Kubernetes, Cadence, Kafka, python, Jupyter/Anaconda, and any other serverless framework that will be identified in the prospective stage.
Target Profile: Service Architect, DevOps Engineer
Serverless architectures refers to the cloud ecosystem that hosts a bunch of services used to compose different end-user applications. The physical host is replaced by an orchestrated container running on docker-kubernetes frameworks.
You will:
24-04-FO / Typescript Architectures for the Multi-Service Front-End
Target M+O: Front Apps Team
Technologies to learn: TypeScript, Javascript, Python, REACT, Swagger, Jupyter/Anaconda.
Target profile: Full-Stack Engineer
One of the biggest challenges for designing orchestrated arquitectures that connect autonomous front-ends with microservices backhauls in the backoffice, is to find a good integration framework that combines synchronous and asynchronous methods, abstracting the service implementations at both sides from the integration complexities and payload.
That is, a simple rich-client app made in REACT should not be aware of the integration detials and loads of the multiple backends to which it connects.
You will:
24-05-APP / Designing Dynamic Low-Emissions Zones for Urban Traffic Scenarios
Target M+O: AI Team
Technologies to learn: python, Keras, Pytorch, Sumo
Target profile: AI Engineer
Focusing on the critical challenge of air pollution in urban areas, primarily caused by vehicular emissions, there is a pressing need for innovative design and management of Low Emission Zones (LEZ) in urban enviroments. Traffic flow callibration in a dynamic way is a critical issue that can be addressed using AI-based methods supported by historical data. This historical data can also be complemented with synthetic data generated with traffic simulations. Once the traffic flows are callibrated, then the traffic planning agency is able to investigate the optimal LEZ, tacking into account not only the emissions volume but also the traffic needs and constraints in a dynamic way.
Besides the preditictive and scheduling capabilities, a dynamic LEZ management system needs to implement the required signalling system, and also the route guidance indications.
You will:
24-06-APP / Distributed Architecture for a Railway Control System
Target M+O: Front Apps Team
Technologies to learn: Python, REACT, Docker, Kubernetes, Google-Cloud, IoT Middleware, Swagger.
Target profile: Full-Stack Engineer
Railway control systems require modern, robust, secure and also flexible applications that can cope not only with the standards but also with the evolutionary requirements that a flexible service imposes.
You will: