CV

Education

• PhD. Candidate in Electrical and Computer Engineering, Yale University, Sept. 2026 Onwards
  • New Haven, CT
  • Working with Professor Boris Landa
• B.S. in Computational Physics with High Honors, University of California San Diego, May 2025
  • GPA: 3.9/4.0, San Diego, CA
  • Relevant Coursework:
    • Upper Division Physics: Classical Mechanics; Quantum Mechanics I & II; Electromagnetism I, II, & III; Statistical & Thermal Physics; Advanced Circuits Lab; Computational Physics II
    • Math & CS: Linear Algebra; Differential Equations; Statistical Methods; Graduate Deep Learning; Data Structures; Machine Learning in Physics; Neuromorphic Computing
• High School, Boston University Academy, Sep. 2017 – June 2021
  • Completed 12 undergraduate courses at Boston University focusing on Physics and Computer Science, Boston, MA

Activities

• Duarte Particle Physics Laboratory, UC San Diego, Jan. 2023 - present, San Diego, CA
  • Developing automated neural architecture search (NAS) and optimization pipeline for deploying efficient machine learning models for low latency hardware like FPGAs
  • Using cutting edge NAS techniques with compression methods like quantization and iterative magnitude pruning
  • Synthesized models onto Vivado FPGA’s for minimum latency model inference
• High Energy Physics Lab (Excellence Research Internship Program), École Polytechnique Fédérale de Lausanne (EPFL), June 2023 - Sep. 2023, Lausanne, Switzerland
  • Created classification program to detect heavy neutral leptons at the CMS experiment at CERN
  • Used Graph Neural networks and multilayer perceptrons to find optimal architecture for signal detection
• HYQU Quantum Physics Laboratory, ETH Zurich, June 2022 - Sep. 2022, Zurich, Switzerland
  • Designed a vibration isolation stage for a cryogenic cavity experiment by deriving predicted eigenmodes and validating with COMSOL; now installed and operating at micro-kelvin temperatures
• MIT Beaver Works Summer Institute, Massachusetts Institute of Technology, Lincoln Laboratories, Jul 2020, (online) Cambridge, MA
  • Developed software packages that integrated processing of satellite images to access damage and analyzed road maps for finding optimal evacuation routes during natural disasters
• A3D3 Trainee, Accelerated AI Algorithms for Data-Driven Discovery Institute, June 2024 – present

Publications/Presentations

• Surrogate Neural Architecture Codesign Package (SNAC-Pack), Dec. 2025, NeurIPS 2025 Machine Learning and the Physical Sciences Workshop
  • Integrated framework that automates the discovery and optimization of neural networks focusing on FPGA deployment without requiring time-intensive synthesis. https://arxiv.org/abs/2512.15998
• wa-hls4ml: Benchmark and Surrogate Models for hls4ml Resource Estimation, Nov. 2025, ACM Transactions on Reconfigurable Technology and Systems (TRETS)
  • Introduced a benchmark and dataset of over 680,000 synthesized neural networks to evaluate GNN- and transformer-based surrogate models predicting FPGA resource usage and latency. https://dl.acm.org/doi/10.1145/3706628.3708827
• Tutorial: Super Neural Architecture Codesign Package (SNAC-Pack), Sep. 2025, Fast Machine Learning for Science Conference 2025, ETH Zurich
  • Led a 1.5-hour hands-on tutorial guiding participants through the complete SNAC-Pack workflow, from dataset preparation to hardware deployment and optimization.
• Neural Architecture Codesign for Fast Physics Applications, Jan. 2025, Machine Learning Science and Technology Journal
  • Automated pipeline to streamline hardware-aware neural architecture codesign for low latency deployment in many scientific fields. https://iopscience.iop.org/article/10.1088/2632-2153/adede1
• Neural Architecture Codesign for Fast Bragg Peak Analysis, Feb. 2024, 2024 AAAI Workshop on AI to Accelerate Science and Engineering, Vancouver, Canada
  • Automated pipeline to streamline neural architecture codesign for fast, real-time Bragg peak analysis in high-energy diffraction microscopy. https://arxiv.org/abs/2312.05978
• Fast ML Conference Lightning Talk & Poster, Oct. 2024, Fast Machine Learning for Science Conference, West Lafayette, Indiana, USA

Awards/Recognitions

• UC San Diego Physical Sciences Dean’s Undergraduate Award for Excellence, Mar 2025, UC San Diego, San Diego, CA
• Sheffield Fellow, Sept 2025, Yale University, New Haven, CT
• Hackathon Winner for “Model Performance” and “Best Judged Model”, Jan 2025, Climate Event Identification NSF HDR ML Challenge Hackathon, San Diego, CA
• Chair’s Challenge Award, Oct 2024, UC San Diego Physics, San Diego, CA
• Physics Honors Program, Apr. 2024 - present, UC San Diego, San Diego, CA
• UCSD Physical Sciences Summer Research Award, June 2024 - Sept. 2024, UC San Diego, San Diego, CA
• Excellence Research Internship Program, June 2023 - Sept. 2023, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland
• Provost Honors, Sep. 2021 - present, UC San Diego, San Diego, CA
• Boston Herald All-Scholastic Athlete, May 2019, Boston Herald, Boston, MA

Skills and Interests

• Computer Languages: Python, Java, C/C++, HTML
• Computer Tools: Tensorflow, Pytorch, Kubernetes, Circuit Building
• ML topics: Computer Vision, Geometric Deep Learning, Fast ML, Network Pruning, Simulated Annealing, Neural Architecture Search
• World Languages: native speaker: English, Russian; B1-level German

Additional Projects

• AI Moving Lamp, Fall 2023, UC San Diego, San Diego, CA
  • DC-powered lamp that moves along a table, uses computer vision to detect notebooks, stops, and turns the light until notebook is no longer detected
  • The lamp uses a Raspberry Pi and a Tensorflow Lite custom trained machine learning model. Designed and 3D printed all the parts of the lamp
  • More info can be seen in this video link
• SQUID Effect Derivation from Maxwell’s Equations, Spring 2024, UC San Diego, San Diego, CA
  • Derived the key equations underlying a Superconducting Quantum Interference Device (SQUID) starting from the most fundamental laws: Maxwell’s and Schrodinger’s Equations
  • Derivations were simplified to an undergraduate level from advanced graduate textbooks that cover this topic
  • The derivation paper can be viewed here (PDF)
• Car Racing Simulation using Reinforcement Learning, Winter 2024, UC San Diego, San Diego, CA
  • Applied Deep-Q Networks (DQN) and Proximal Policy Optimization (PPO) models to optimize performance in the OpenAI Car Racing game environment
  • Explored reward augmentation strategies like grass detection, speed rewards, and acceleration rewards to enhance training efficiency and model performance
  • View the presentation and detailed paper