Designed and implemented a robust framework to allow for backwards compatible MongoDB database migration for our software upgrade, making use of systemd, docker containers, and pymongo library.

Designed and implemented the backend stack for providing the network ACL feature, utilizing linux iptables, docker, and python celery/bottle.

Worked directly with customers to resolve issues on-demand, such as providing a timely patch to HSBC to provide them a backdoor access to cleaning up InfluxDB with corrupted metadata.

Overhauled build/deployment stack for the product various times, such as providing the stack the ability to run with dynamic configuration for cache/memory limits for celery and influxdb based on hardware.

Enhanced logging and debugging interface for developers and customers with enhanced logging by setting up rsyslog server/logrotate to manage and preserve logs, redirecting docker logs to the rsyslog server.

Worked with the pandas and polars libraries to develop and improve the network packet analytics engine and set up resource control to avoid running out of memory and bottlenecking the InfluxDB.

Self studied Monte Carlo methods/theory then implemented an Ising and Heisenberg square lattice Monte Carlo model with C and Python and verified by computing theoretical results.

Implemented a Honeycomb Heisenberg model under symmetric exchange and Dzyaloshinskii-Moriya (DM) interaction.

Performed analysis of resultant spin configurations. Discovered degeneracy in the vortex-like ground state.

Redesigned Malloc Lab Baseline for the students and wrote a simulator/parser to show students how allocation/freeing interacted with coalescing/etc.

5th most contributions on Piazza out of 40+ TAs, with ~500 personal unique responses.

Rewrote Malloc Lab handout and recitation slides, made general posts regarding concepts from class and for addressing misconceptions, worked many hours in office hours to help students, and answered student questions outside of work schedule.

Redesigned framework for 20-year-old webpages to present experimental crystal data dynamically based on alloy database.

Built and deployed the Python backend in Django and Apache (both self studied on the job), and successfully utilized a geometrical theory developed with the professor to generalize to infinite dimensional crystals to calculate energy offsets from the alloy’s convex hull.

Utilized Highcharts API to make binary and ternary plots for convex hulls of alloys in javascript, and JsMOL to display interactable 3d crystals in putting together the frontend.

Developed Python code to extract data from Raspberry PI in real time and generate spectrum plots with matplotlib.

Constructed UI to calibrate the collected data, so the spectrum data could be compared to the theoretical model of the reionization period of the Universe’s early stages.

Travelled to Green Bank, West Virginia with the professor to conduct the experiment, using my program to monitor the experiment and analyze the results.