Marek S. Email and Phone Number

• Designing and implementing end-to-end solutions for speech recognition (both offline and real-time) including training deep neural network models• Designing, implementing, and training deep neural network models• Interpretation of the results of introducing corrections/improvements to the model or training• Analyzing data• Integrating with other systems• Active research in the area of ​​speech recognition• Research in the area of LLMs

Note: Noble Systems company merged into Alvaria• 02.2020 - 08.2020: C++ backend service• 09.2020 - 06.2021: Frontend⠀◦ Typescript, React, WebRTC⠀◦ Technically leading project

I have worked on a series of projects using mainly C++ (11, 14, 17) and the Qt framework. My role involved being responsible for the technical aspects of the projects. This included designing the software from scratch, maintaining it, and ensuring its reliability and correctness through unit and integration testing. In some projects, I was also responsible for establishing technical requirements with the client.Projects:• Android app for opticians to measure distances e.g., between pupils, to accurately select lenses for patients. (OpenCv used for vision processing)• Industrial Machine Control Application for a paper perforation machine (Communication with the machine PLC via the OPC-UA protocol)• Versatile application for generating document templates based on user inputs

• Designing and developing the new cryptocurrency (forked from another cryptocurrency)• Adapting other cryptocurrency-related software(pools, miners, wallets, etc.)• Deploying the core components of the cryptocurrency software ecosystem andstart network• Testing• Debugging• Providing support and consultancy concerning blockchain technology

• Developing a component of a WCDMA base transceiver station• Implementation of new functionality within that component⠀◦ Among others, Dual/multi-carrier HSDPA/HSUPA• Maintenance and refactoring of existing code• Integration and unit testing (GTest)• Working in a team of 7-9 people, using SCRUM methodology