Foad Jafarinejad Email & Phone Number

Munich, BY, DE

Munich, Bavaria, Germany

- Leveraged industry best practices: Conducted in-depth research on Android cryptography APIs, ensuring alignment with NIST and OWASP recommendations for robust mobile security.- Streamlined build processes: Deepened expertise in Gradle build system through workshops and implemented best practices, optimizing cache management and significantly reducing.

Voluntary Online Community

- Collaboratively reviewed mobile application security risks with the team, identifying potential vulnerabilities.- Contributed to the analysis of risk impact, exploitability, and detectability of identified vulnerabilities.- Reviewed draft versions to ensure clear and concise communication of security risks, catering to both technical and non-technical.

Heidelberg, Baden-Württemberg, Germany

- Conducted a deep dive into Solana's consensus protocol through comprehensive documentation review and source code analysis. This identified key validator functionalities.- Designed and executed controlled experiments simulating the Solana consensus protocol. Analyzed the impact of network delays on the protocol's liveness, ensuring its ability to.

Darmstadt, Hesse, Germany

- Developed a program to automate the retrieval of third-party libraries from mvnrepository, facilitating efficient analysis.- Proposed a novel algorithm for deobfuscation, employing method size and unsupervised classification (k-means) to reliably identify classes within obfuscated applications.- This project contributes to improved security and.

Darmstadt, Hesse, Germany

- Identified a Gap: Recognized the need for better debugging tools in deep learning programs.- Proposed an Abstraction-Based Approach: Presented a novel solution to detect deep learning bugs using an intermediate neural network representation.- Implemented the intermediate representation and tested on sample code.

Analyzing Disinformation on Twitter by - Identified users' relationships in a political campaign using causal inference. - Defined time series analysis to establish user behavior causality.

- Provided Hands on Deep Learning lectures.- Designed and graded captivating exercises and projects.

Remedying High Dimensional Causal Structure Learning by - Reduced error propagation in skeleton learning of existing algorithms. - Developed order-independent PC algorithm for high-dimensional causal structure learning.

Head teaching assistant of Data structure and Algorithms, providing leadership and support for fellow TAs and graders to ensure smooth operation.

Developed "cuPC," a GPU-based parallel algorithm for learning causal relationships from high-dimensional data. This significantly outperforms the state-of-the-art (1300x faster execution with improved accuracy).

Lecturer of Advanced Java Programming. Promote student success by designed engaging materials, exercises, and projects.

Head Teaching Assistant of Data Structure and Algorithms, providing leadership and support for fellow TAs.

Implemented and trained DeepWarp model for gaze manipulation on the Columbia Gaze estimation dataset.

Personalized Arrhythmia Detection for Wearables: Developed a LSTM-based algorithm for Android wearables to personalize the detection of heart rhythm abnormalities (arrhythmias) from ECG signals.

Master'S Degree, Computer Science

Activities and Societies: Research Assistant

Bachelor'S Degree, Computer Science