Competition for talented FI students - April 2025

The Association of Industrial Partners of the Faculty of Informatics of Masaryk University (SPP FI MU) in cooperation with Lexical Computing, Progress Software, Monet+, Red Hat Czech, Y Soft and Honeywell
Competition for talented students of FI 2025.

The competition is open to students in the 2nd and 4th semesters of undergraduate studies. The competition will include several tasks in computer science, which will take 24 hours to solve. The final part of the competition will take place in the form of interviews with representatives of FI laboratories and partner companies. The finalists can get one of the student researcher positions with a scholarship.

Stipends will range from 4-8k per month, depending on the intensity of involvement in the lab. A specific supervisor from the staff or Ph.D. student's lab will be identified for each position.

Competition process (details to be added):

• Monday, April 7, 2025 at 9:30 - first set of tasks to be download in IS.
• Tuesday, April 8, 2025 at 9:30 - second set of tasks to download
• Thursday, April 10, 2025, 14:00 - 16:00 in room S108 (building S) - personal interviews with the finalists of the competition will take place and specific students will be selected for the laboratories

Competitors will be informed by email no later than Thursday, April 10, 2025 at 9:00 whether they have advanced to the finals of the competition.

Lab positions offered:

Natural Language Processing Lab is looking for 1 student to develop and conduct research in numerical vector computation used in training and inference of large language models. The goal of the research is to speed up both tasks, especially in a CPU environment, which is beneficial for making larger models accessible to a wider range of users with common hardware.
Start date: immediately or by appointment
Sponsor: Lexical Computing

Natural Language Processing Lab is looking for 1 student to develop tools for efficient management of text corpora in a distributed cloud environment. The cloud environment is based on Linux distributions and includes higher tens of nodes with distributed storage accessed by users (developers) and automated tools. The goal of the development is to develop support systems to manage and maintain the cloud and its storage.
Start date: immediately or by appointment
Sponsor: Lexical Computing

Natural Language Processing Lab is seeking 1 student to develop tools for efficient parallel processing of text corpora in a distributed cloud environment. The cloud environment is based on Linux distributions and includes higher tens of nodes with distributed storage accessed by users (developers) and automated tools. The goal of the development is to develop tools for massively parallel processing of text corpora across distributed storage.
Start date: immediately or by appointment
Sponsor: Lexical Computing

Cybersecurity Laboratory is looking for 1 student to conduct research and development in the area of network security with a focus on detecting network service problems through time series analysis. The goal is to create a tool that will analyze network traffic data and identify problems (outages, unusually high traffic) of a particular network service. The development will be done in Python and relevant technologies for data processing and analysis.
Start date: After the end of the test period (or by appointment)
Sponzor: Progress Software

Centre for Research on Cryptography and Security (CRoCS) is looking for 1 student to develop the SCRUTINY tool (https://github.com/crocs-muni/scrutiny) for comparing the characteristics of cryptographic library and security device type security products. The target product is first subjected to a set of tests (e.g. supported algorithms, speed, sampling during the execution of a cryptographic operation...), the results of which are stored for later comparison as a forensic template. Repeated measurements over a (presumably) similar product are compared against the prototype and evaluated to detect changes in internal implementation, different hardware or limitations in functionality. The primary focus of development within this position is on developing a tool that compares and visualizes the output of domain-specific sub-tools. The core tool is implemented in Python, with integration of inputs with a variety of other tools. We are interested in a longer-term collaboration that may naturally transition into a bachelor's thesis solution or internship at Monet+.
Start date: immediately or by agreement
Sponzor: Monet+

Centre for Research on Cryptography and Security (CRoCS) is looking for 1 student to develop tools for analyzing cryptographic implementations on software libraries and cryptographic chips. Depending on the specific focus, the tools sequentially call cryptographic functions and process their result including performance metadata. In contrast to the previous position (Monet+ & SCRUTINY), in this case it directly extends the sub-tools listed in this table - for example jcalgtest, tpmalgtest, jcAIDScan or ec-detector. These tools are typically implemented in C/C++, Java or Python. We are interested in a longer-term collaboration, which may naturally transition to a bachelor thesis or internship at Monet+.
Start date: immediately or by agreement
Sponzor: Monet+

Centre for Research on Cryptography and Security (CRoCS) is looking for 1 student to develop and conduct research in the area of linking security-relevant metadata (result of randomness tests, side channels, implementation weaknesses...) and certified security and cryptographic devices, systems and libraries. The goal is to extend the system, populate it with additional data and allow users to easily obtain relevant information - all using open data formats and a decentralized network (Nostr). In addition, the sec-certs.org web application with security certificates will be extended with a module that analyses and displays this metadata in a suitable way. The development is being done using common web technologies and Python. We are interested in a longer term collaboration that may naturally transition into a bachelor thesis solution.
Start date: immediately or by agreement
Sponzor: Red Hat Czech

Centre for Research on Cryptography and Security (CRoCS) is looking for 1 student to conduct research and development in the area of data analysis of security device certificate documents (Common Criteria, FIPS 140) and further development of the https://seccerts.org tool . The first step will be to become familiar with the available data and the certification ecosystem. The main goal of the collaboration will then be to extend the existing data analysis and visualization. Depending on the student's interests, the work may be more implementation or more analysis work. The work will be done in a team-based manner in collaboration with the company and other lab members. We expect an interest in applied security and knowledge of Python (no need to be an advanced programmer or have prior knowledge of certifications). We are interested in a longer-term collaboration that may naturally transition to a bachelor's thesis solution.
Start date: immediately or by agreement
Sponzor: Red Hat Czech

QualiFI lab in collaboration with Red Hat is looking for a talented student to collaborate on tool development and research in the area of program performance analysis. Are you interested in operating systems, profiling, performance analysis of user programs and/or the Linux operating system kernel, data visualization, advanced algorithms and data structures for performance analysis or issues related to (semi-)automatic detection of performance changes? Do you want to contribute to research in any of these areas and to the development of Perun, a tool used at Red Hat to analyze performance changes across Linux kernel versions? You don't need to be an advanced programmer or have extensive knowledge in performance analysis; analytical thinking, an interest in C/C++/Python programming (preferred) and a desire to find innovative solutions to complex problems are key. In case of mutual interest, we offer the possibility of long-term collaboration, which may continue in the form of bachelor, master or PhD theses.
Start date: immediately or by agreement
Sponzor: Red Hat Czech

QualiFI lab in cooperation with Red Hat is looking for a talented student to collaborate on the development of tools and research in the field of static program analysis to automatically detect unwanted changes in programs. Specifically, this is a collaboration on research and development related to the DiffKemp tool, which is designed to automatically detect when a modification to a program that was intended to make it, for example, cleaner or more efficient, but not intended to change its function, accidentally changes that function. The emphasis is on the tool being able to handle large, complex, low-level optimised projects such as the Linux kernel. Project work is varied and can include, depending on individual preferences, programming, experimentation and inventing new approaches with deeper formal roots, even changing over time. However, it is not necessary to have an immediate knowledge of advanced programming, operating systems, or the mathematical foundations of static analysis. What is important is an interest in the field, a willingness to try something at the interface of programming and methods with formal foundations, and a desire to analyze and solve complex problems in order to push oneself. We are interested in a long-term collaboration, which may develop into a bachelor's thesis, a master's thesis, and possibly later towards a PhD.
Start date: by appointment
Sponzor: Red Hat Czech

Embedded Systems Laboratory (EmLab) is looking for a student for research in the field of artificial intelligence and computer vision. The student will participate in the design and development of a virtual smart robot that uses a combination of traditional computer vision algorithms and deep learning models to interact with the virtual device. The research and development includes extending the capabilities and making the virtual robot work with different types of devices (Websites, Desktops, Mobile, ..). The prerequisites are a willingness to program in C# or Python, an interest in the basics of neural networks and computer vision, and a willingness to learn. We offer the possibility of long-term cooperation in the field of artificial intelligence and computer vision with the assumption of collaboration on the final thesis. In case of mutual interest, it is possible to further cooperate with Y Soft.
Start date: start immediately or by agreement
Sponsor: Y Soft

QualiFI lab, in collaboration with Honeywell, is seeking a talented student to collaborate on the development of tools and research in the analysis of specifications and models of aerospace (or other safety-critical) systems to automatically detect faults and generate tests for certification purposes. In the case of specifications, this specifically involves the validation of informal or semi-formal specifications of safety-critical systems, such as aerospace components or medical devices, using logical reasoning or other formal methods such as SMT solving. For these systems, the emphasis is on correct behavior under all possible conditions and on detecting any problems at the earliest possible stages of development, when correcting these problems is still relatively inexpensive. In addition to specification validation, the focus can also be on their formalization, e.g. using LLM models and other modern approaches in the field of generative artificial intelligence. For models, this mainly involves the analysis of models created using specialized avionics libraries for MATLAB/Simulink, such as the BLADE/BALROG public libraries developed in collaboration with the research groups QualiFI FI MU, VeriFIT FIT BUT and Honeywell. Current analyses focus on automatic test generation for certification purposes, which are highly demanding in terms of both requirements and code coverage, using SMT solvers such as Z3. However, other approaches or analysis focuses are also under consideration. Familiarity with MATLAB/Simulink is not necessary as all analyses run over the internal model representation provided by the BLADE/BALROG libraries. Work in this area may include research and development of new analysis algorithms, implementation, experimentation with them, or research into the actual design of safety-critical systems to make them as simple to specify as possible and as easy to analyze as possible. An interesting problem is also the presentation of an understandable explanation of the analysis results in a form acceptable to the certification authority as evidence of compliance with the requirements imposed on the development of safety-critical systems. To start working in this area, it is not necessary to have a deep knowledge of safety-critical systems analysis methods, their specifications or MATLAB/Simulink models. What is important is an interest in the area, including the mathematical roots of the methods used, and a desire to solve complex problems at the interface between formal methods and advanced algorithmization. We are interested in a long-term collaboration, which may evolve into a bachelor or master thesis, and eventually lead to a PhD.
Start date: by appointment
Sponsor: Honeywell

QualiFI lab, in collaboration with Honeywell, is looking for a talented student to collaborate on the development of tools and research in the area of static analysis of control programs of aircraft (or other safety-critical) systems using artificial intelligence in the form of expert systems. Specifically, these are expert systems based on logical rules that should either assist pilots in solving complex situations or directly decide on actions performed by autonomous aircraft systems. Of course, the properties of such systems need to be carefully analysed, including the unambiguity of the decision, the finiteness of the reasoning or the consumption of the computational resources needed to do so. All of the above is a largely open research challenge, which is the subject of joint research by QualiFI Laboratory, Honeywell Advanced Technologies and some of their other partners. Work in this area may include research and development of new algorithms for analysis, implementation, experimentation with them, or research into the expert systems themselves to make them as efficient as possible at runtime, but also as analyzable as possible. Giving an understandable explanation of how a decision was reached is also an interesting problem. It is not necessary to have a deep knowledge of methods for analyzing computer systems, expert systems, or safety-critical systems to get started in the field. What is important is an interest in the area, including the mathematical roots of the methods used, and a desire to solve complex problems at the interface between formal methods and advanced algorithmization. We are interested in long-term collaboration that may develop into a bachelor's thesis, a master's thesis, and possibly later on towards a PhD.
Start date: by appointment
Sponsor: Honeywell