Untitled
unknown
plain_text
a year ago
3.9 kB
7
Indexable
## EXPERIENCE <span style="font-size: 1.2em; font-weight: bold;">Software Engineer</span> Samsung RnD Vietnam <span class="iconify" data-icon="tabler:map-pin"></span> Hanoi, Vietnam ~ <span class="iconify" data-icon="tabler:calendar-week"></span>06/2024 - Present System Security - Implement and improve features of the Android Security System. - Participate in design, implementing and maintain the code-based within Security Team. <span style="font-size: 1.2em; font-weight: bold;">Software Engineer Intern</span> Samsung RnD Vietnam <span class="iconify" data-icon="tabler:map-pin"></span> Hanoi, Vietnam ~ <span class="iconify" data-icon="tabler:calendar-week"></span>03/2024 - 06/2024 Secure Vault - Developed a secure vault for storing and delivering sensitive data within Android applications. I built a storage vault allowing applications to securely store sensitive data such as tokens and keys. Other applications can send API requests to my service to store, request, modify, or delete data. The project consists of three main modules: register, storage, and query. I developed the workflows, core functionalities of the service, and managed the team of four members. I was responsible for the register module, which allows other apps to register for the service. The app layer was built using Java, while the native layer was implemented with C++ using OpenSSL. - This project was awarded the Best Project by the Security Council. <span style="font-size: 1.2em; font-weight: bold;">Research Assistant</span> Hanoi University of Science and Technology (Hanoi, Vietnam) <span class="iconify" data-icon="tabler:map-pin"></span> Hanoi, Vietnam ~ <span class="iconify" data-icon="tabler:calendar-week"></span>11/2023 - Present Quantum Neural Network - I built a small QNN library that includes classes such as Feedforward, CNN, and LSTM. Important formulas for training networks, like gradient calculations, were researched beforehand (in quantum computing, the shift parameter rule is used for derivatives). The network structures are similar to classical neural networks, and the classes I built can be combined with classical networks to form hybrid networks. Since I did not have access to a quantum computer, I used a simulator, resulting in good outcomes due to the absence of noise. - I applied Transfer Learning in QNN to classify MNIST images, using methods similar to classical approaches. - I proposed the Quantum Quantization Aware Training (QQAT) method to enhance the efficiency of QNNs on resource-constrained devices. MageCode - a code analysis system for the Bach Khoa Cyber Security Center: - This project was submitted for the SVNCKH competition. It consists of three modules: analyzing whether the code was written by a human or machine, checking for code plagiarism if written by a human, and checking for logical errors such as array overflows and division by zero. I was responsible for the third module, using LLM for error checking and CodeQL from GitHub for static analysis. This project is still under development at the 405th floor of the Institute of Information and Communication Technology to become a full product. I served as the PM because I proposed the idea (with support from professors), managed the team, and designed the main workflows (determining which module the source code should enter first, which tools to use for analysis, how to conduct checks, and identifying the target audience for the product). <span style="font-size: 1.2em; font-weight: bold;">Software Engineer Intern</span> Viettel Telecom <span class="iconify" data-icon="tabler:map-pin"></span> Hanoi, Vietnam ~ <span class="iconify" data-icon="tabler:calendar-week"></span>08/2023 - 11/2023 - Analyzed VESS (Viettel Enterprise Support System) developed by the company in Odoo Python and produced tenical documentation. - Implemented some module using Odoo.
Editor is loading...
Leave a Comment