Hi! I'm a senior at UC Santa Barbara studying computer science. I'm currently a master's student and I'll be interning at the Lawrence Berkeley National Lab this Summer. I'll be working on computer networking (traffic classification) research.
I have a certificate I received after taking courses covering Linux, network security, and TCP/IP. When I interned at Axiado I developed a Python package that would take packet flow descriptors as input, and output these flows both as raw bytes to be used in simulation, or as packets that could be sent over the wire. The following Summer, I used this tool to test a firewall's packet parser. This testing not only involved creating network traffic the chip would encounter during normal operation, but also packets with exotic headers, and even simulated cyberattacks! At UCSB, I've taken an undergraduate computer security course and a graduate course covering classic papers in the field. I'm currently taking a graduate security seminar, and I'm also working on IDS research in our Computer Security Group.
At Cadence Design Systems, I developed an automation tool to create memory wrappers, i.e. combine smaller foundry-provided memories into larger ones. I also have experience with programming in CUDA C++ for the Nvidia A100 from when I worked at OpenEye Scientific. During my internship at Arista Networks, I designed a C++ data structure to decode registers in Broadcom's switching ASICs. I read the chip's technical specifications, and architected a system that could translate the information encoded in these registers to a format readable by Linux processes that needed to act on it. At UCSB, I've taken a computer architecture class that covered cache designs, paging, and branch prediction and I'll be taking a graduate computer architecture course this Fall.
Last year, I worked under Professor Arpit Gupta on applying machine learning to computer networks. We developed an LLM agent to diagnose network issues (e.g. low signal strength) and suggest a solution (e.g. moving closer to the wireless access point). My contributions were building a database for this tool to make it RAG-capable and developing network measurement pipelines. We published our results on arXiv here. I've previously taken courses in graph neural networks, machine learning for networking, artificial intelligence, and optimization and I'm currently taking courses on foundation models and frontier LLM and agent capabilities.
You can find my resume here.
