Resume
Professional Summary
PhD candidate in Mathematics at UT Austin with industry experience in LLM fine-tuning and agentic AI systems at AMD. Research interests include ML for scientific computing, nonlocal PDEs, mathematics of LLMs, and AI for mathematics. Seeking research scientist, ML/AI engineer, and software engineer positions.
Technical Skills
- Languages & Libraries
- Python (PyTorch, NumPy, JAX, scikit-learn, SymEngine), C/C++
- Tools
- Git, Linux, Bash
Experience
- Aug. 2024 - Dec. 2024
- GPU/ML Researcher/Developer Intern (part time), Advanced Micro Devices (AMD)
- Integrated fine-tuned large language models (Qwen, DeepSeek, Llama) into an agentic AI framework to automate Fortran-to-HIP code translation.
- Designed a multi-agent system using CrewAI, where coordinated agents generated, reviewed, and debugged translated GPU code, achieving higher translation accuracy and performance compared to the zero shot finetuning baseline.
- May 2024 - Aug. 2024
- Co-Op/Intern: PhD Tech, Advanced Micro Devices (AMD)
- Fine-tuned open-source LLMs (Qwen, DeepSeek, Llama) to automate Fortran-to-HIP code translation, enabling the generation of a first draft GPU compatible code.
- Built and curated a dataset, used in finetuning, of input and output pairs that correspond to Fortran code and the equivalent parallel code in HIP.
- Jun. 2023 - Present
- Graduate Research Assistant, Oden Institute, UT Austin
- Designed and deployed a high-speed synthetic data generation pipeline for neural operator training, accelerating PDE solutions generation for scientific ML models.
- Developed a Python library leveraging SymEngine to generate PDE datasets used in training neural operator architectures.
Education
- Aug. 2021 - present
- University of Texas at Austin (Austin, TX)
PhD Candidate in Mathematics (expected graduation: May 2026)
Research focus: PDEs in operator learning and non-local operators.
Advisors: Prof. Stefania Patrizi and Prof. Rachel Ward - Jan. 2020 - May 2021
- Florida Institute of Technology (Melbourne, FL)
M.S. in Applied Mathematics.
Thesis Title: “The Brezis-Nirenberg Problem for the Generalized Kirchhoff Equation” - Aug. 2016 - Dec. 2019
- Florida Institute of Technology (Melbourne, FL)
B.S. in Applied Mathematics
Publications
- 2026
- The strongly nonlocal Allen–Cahn problem. Erisa Hasani, Stefania Patrizi. Journal of Differential Equations, Volume 462. arXiv
- 2025
- Generating synthetic data for neural operators. Erisa Hasani, Rachel A. Ward. The SMAI Journal of computational mathematics, Volume 11 (2025), pp. 497-516. arXiv
- 2023
- On the critical \(p\)-Kirchhoff equation. Erisa Hasani, Kanishka Perera. Topological Methods in Nonlinear Analysis Vol. 61, Issue 1, (Jan 2023), pgs 383-391. arXiv
- 2021
- On the compactness threshold in the critical Kirchhoff equation. Erisa Hasani, Kanishka Perera. Journal of Discrete and Continuous Dynamical Systems, 2022. arXiv
Honors
- May 2022
- Frank Gerth III Teaching Excellence Award, University of Texas at Austin
- Feb. 2013
- Bronze Medal, 30th Greek Mathematical Olympiad (Athens, Greece)