MIT-Spain

Universidade da Coruña (University of a Coruña) - UDC

Centro de Investigacións en Tecnoloxías da Información e as Comunicacións (Center for Research in Information and Communications Technologies) - CITIC

Are you interested in working at the frontier of Artificial Intelligence (AI) and Cybersecurity? CITIC Research Center, located in the scenic coastal city of A Coruña, Spain, invites MIT students to experience an exciting two-month internship. We want you to join us in the development of cybersecurity solutions based on the base technology of applications such as ChatGPT or Gemini, the Large Language Models (LLMs).

One of our most promising research lines is focused on addressing one of the most critical challenges of modern telecommunication networks: ensuring security in increasingly complex environments, such as those involving 5G/6G or Internet of Things (IoT). With the rapid growth of digital infrastructures, cyber threats are also evolving, and traditional AI models are struggling to keep up. Current models are typically designed for specific tasks and fail to generalize well across diverse network environments, making them vulnerable to previously unseen threats.

Our Aim:
At CITIC, we aim to develop foundational AI models capable of generalizing across a wide range of network environments and tasks, bringing much-needed flexibility and robustness to cybersecurity. The potential of foundational models—AI systems that can be fine-tuned to perform well in multiple environments without retraining from scratch—promises to revolutionize the way networks are protected.

We are leveraging the Transformer architecture, which has already demonstrated extraordinary success in Natural Language Processing (NLP) and Computer Vision (CV), with models like OpenAI o1 or GPT-4o, to create network traffic models. These models will be able to analyze vast amounts of network data, detect security anomalies, and predict potential cyber threats in real time. By doing so, we are pushing the boundaries of AI applications in cybersecurity, building more efficient, autonomous solutions to protect networks. This research aligns with the global AI trend of developing general-purpose AI models capable of addressing challenges across diverse fields and environments.

Your Role:
As an intern, you will be immersed in the development and fine-tuning of these foundational AI models. You will have the opportunity to:
● Work closely with our team of AI and cybersecurity experts to develop and optimize Transformer-based models for the analysis of network traffic data.
● Participate in cutting-edge research exploring how AI models can generalize across diverse network environments, such as data centers, 5G networks, and IoT systems.
● Contribute to real-world applications of AI in network security, including anomaly detection, traffic pattern analysis, and the classification of malicious network activity.

You will be actively involved in the model development pipeline, from data acquisition to model training and evaluation. This will involve handling vast amounts of unlabeled network data, pre-training models to learn from this data, and then fine-tuning them for specific tasks like threat detection or malware identification.

What makes this project particularly exciting is its potential impact on the future of AI in cybersecurity. Foundational AI models promise significant efficiency gains in network management, reducing the need for building task-specific models. Moreover, we can enhance the scalability of our AI models, making them faster to train and easier to implement across different network environments. This combination of generalization, efficiency, and adaptability will be key to shaping the next generation of network security solutions.

As part of this internship, you will join RNASA-IMEDIR, one of the largest and most accomplished research groups at the University of A Coruña (UDC), led by Dr. Alejandro Pazos Sierra, a renowned expert in AI. The group has a rich history of innovation in AI across various fields, including cybersecurity, biomedicine, and digital transformation, and holds the majority of teaching hours in cybersecurity at the university. The group actively participates in European projects, with a current example being EDINAF, where we contribute to advancing AI-driven digital transformation while addressing key ethical considerations. With a strong international presence, numerous publications, patents, and the successful creation of AI spin-offs, the group provides a dynamic and cutting-edge environment for advancing research, making it the perfect setting for any student eager to explore AI-driven solutions in network security."

Galenicum

Galenicum is a dynamic pharmaceutical company created 20 years ago by three friends in their 20s (now 40s). The three friends remain still friends, and are very much involved in the business as co-CEOs of the companies within the Galenicum Group (B2B and B2C businesses). The company keeps a remarkable entrepreneurial spirit and energy, with young staff from both scientific backgrounds in Life Sciences, as well as business profiles and Engineers.  

The values of the company are quite embedded in the culture, in fact the culture was the origin of the values exercise, and they are:

Challenge - entrepreneurial, super dynamic
Connection - good relationships and interconnections within people, providers, clients - WFH 1 time a week only, lots of people in the office
Commitment - engaged with the company and focussed on results

Booster values are: Energy, Adaptability, Creativity

Integrity - critical at all levels

Galenicum has been working on ESG over the past years, both motivated by genuine belief and by the business and compliance needs. We have articulated a Strategy over the past 3 years that we are happy with, but we want to go beyond and understand what shall be next for Galenicum in ESG, understanding well the right benchmarks and remarkable initiatives in the field and recommending what shall be done in the next 3-5 years.

HM Hospitales

Comprehensive Cancer Center CLARA CAMPAL, Hyper-Early Cancer Detection Unit

Madrid, Spain

At a prestigious cancer center in Spain, university students rotate through a specialized Hyper-Early Cancer Detection Unit, where they gain valuable experience in both cutting-edge diagnostics and research through clinical trials. This unit is dedicated to detecting cancer at its earliest, often pre-symptomatic stages, using innovative techniques like liquid biopsies, advanced imaging, and genetic profiling. Students work alongside a multidisciplinary team of oncologists, radiologists, and researchers, observing how these advanced methods significantly improve early diagnosis and patient outcomes.

During the rotation, students are involved in patient assessments, learning to apply hyper-precise diagnostic tools that detect cancer in its most nascent stages. They also participate in clinical trials, where novel detection methods and treatments are tested, contributing to the research process that drives oncology forward.

This hands-on experience provides a deep understanding of both clinical practice and the critical role research plays in advancing early cancer detection. By combining patient care with innovative research, the program prepares future healthcare professionals to play a vital role in the evolving fight against cancer.

Barcelona Supercomputing Center

Data Analytics and Visualization Group

The project will involve the student(s) with the preparation and creation of two related installations to be set up in the Design Hub Museum in Barcelona. The first one is a reconstruction of the Cybersyn (https://en.wikipedia.org/wiki/Project_Cybersyn), a connected control room to make informed decision-making created in the early 70s. The second one is an installation about the Barcelona Digital Twin, an ongoing project to create a digital replica of the city to allow urban planners and managers to visualize the impact of potential policies (https://www.vcity.tech). Separated by 50 years of development, it's not just faster and better technology, because now not only the people in the room can make decisions, but also every citizen connected to the platform can see and participate in the process. 

The student project will start by working with Eden Medina, creator of the Cybersyn reconstruction displayed in Chile, and BSC helping set it up and optimize the technical equipment. Then, the students will participate in the creation of the Digital Twin installation, which is currently being conceptualised.

The students will familiarize themselved with the Cybersyn reconstruction and technology (based in microcomputers) and help set up the installation in Barcelona Design Museum. Afterwards, the students will work with the BSC team to help create the installation of the Barcelona Digital Twin that will be displayed opposite the Cybersyn.

Universidad Nacional de Educación a Distancia (UNED)

Departamento de Física Interdisciplinar

The Laboratory of Infrared and Raman Spectroscopy (https://portalcientifico.uned.es/grupos/17451/detalle) conducts experimental and theoretical studies in polymer materials, and atmospheric and astrophysics systems. Our labs and our offices are located in Las Rozas de Madrid, which is a 30-40 minute bus ride from the Moncloa metro stop in Madrid, which is really close to the city.

Our lab is composed of 5 members with backgrounds in Physics, Chemistry, and Materials Science. We go to work to our university labs and offices every day of the week, although our schedules are flexible. All of us are young people that are active in sports (cycling, running, padel) and like to spend time outdoors. 

Our main project explores the use of photocatalysis to the degradation of micro and nanoplastics. Plastic is, nowadays, a ubiquitous material in our daily life, with global production close to 370 million tons. As everybody know, its environmental impact is a significant concern, as only a small fraction is recycled. The plastics accumulated in the environment can be fragmented, resulting in the so-called microplastics (MPs, <5 mm) and nanoplastics (NPlast, <1 μm). Effective elimination of plastic waste from the environment is a pressing research goal, but plastics are difficult to degrade due to their low water solubility, complex structure and non-biodegradability. Current methods are not enough to solve this question, making the exploration of alternatives essential. Photocatalytic technology has emerged as a promising solution to degrade polymeric materials, with a focus on MPs and NPlast. In this topic, we have available student projects in areas 5, 8 and 10 to explore the role of photocatalysis for the plastic degradation. We will generate highly porous nanostructures of different catalysts to degrade commercial plastics analogues produced in our labs. Our basic studies, both experimental and theoretical, allow us to extract fundamental conclusions of these processes, and to acquire valuable insights for further advancements in this field.

The student will conduct experiments (under supervisor) to produce nanoporous photocatalysts and will tests their activity on the degradation of thin films of usual polymers.

Universidad Politécnica de Madrid

Department of Electrical Engineering, Electronics, Automation and Applied Physics 

Our team at ETSIDI, focused on the research group topic Low and high voltage networks and installations, fosters a culture of innovation, collaboration, and excellence. We are a diverse group of professionals and students who are passionate about electrical engineering. Our office environment is dynamic and supportive, encouraging creativity and continuous learning. We value teamwork and believe that the best solutions come from sharing ideas and working together. With a strong emphasis on real-world applications and industry partnerships, our culture is geared towards preparing students for successful careers and advancing the field of electrical engineering. We have extensive experience in participating in research projects and in international collaboration.

Over the past decade, retrofilling of mineral oil-filled transformers with natural or synthetic esters has expanded significantly. The main reasons for retrofilling a transformer are to reduce the fire and environmental risk of the equipment, as well as to increase its load carrying capacity and service life. Natural and synthetic esters are biodegradable materials and have much higher flash points than mineral oil. Furthermore, the use of esters has been shown to reduce the aging rate of the transformer’s solid insulation. Since the long-term overload capability of a transformer is limited primarily by the aging rate of its solid insulation, retrofilled transformers can operate at higher loads without decreasing their service life. In this project, the impact of retrofilling on the transformer’s dielectric behavior and the impact of retrofilling on the service life of the solid insulation are investigated. In addition, aspects related to the maintenance and operation of retrofilled transformers are investigated.

This project is funded by the Spanish Government and it is a coordinated project in which several institutions and companies are involved on.

The next task will be developed by the group led by Ricardo Albarracín Sánchez.

Modelling of electric field distribution in retrofilled transformers
The electric field distribution of the transformer representative geometries selected in this task will be analysed using Finite Element Method (FEM) and considering different voltages waveforms (i.e., AC, lightning impulse full-wave and impulse chopped-wave). The dielectric margins of the different parts of the insulation will be calculated and compared with the dielectric margins of the transformer insulation before retrofilling.

Developing a guide on the need for transformer derating or design review.
The conclusions extracted from previous tasks will be compiled to produce guidance on whether a transformer design must be re-evaluated to assure the safety of its retrofilling with an alternative insulating fluid. The guide will be presented to actors involved in transformer retrofilling.