Hi! I’m Rishmita
Problem solving enthusiast looking to bring my multifaceted skills to the tech world
About Me
I take every opportunity I possibly can to explore the expansive universe of software development and scientific programming. I love learning new technologies to push the limits of my passions. I'm currently looking for entry level positions in software engineering, forward deployed engineering, or data engineering.
I can talk forever about astronomy, music, sci-fi (particularily sci-fi short stories!), synthesizers, and daily games (especially if you're a minutecryptic player). My hobbies are ever changing and always shifiting but I think that makes life more fun :D
Skills
Programming Languages
Python · C/C++ · JavaScript/TypeScript · HTML/CSS · SQL
Fullstack Frameworks
React.js · Express.js · FastAPI · Jekyll
Data Science Frameworks
Pandas · NumPy · SciPy · Seaborn · Streamlit
ML/AI Frameworks
TensorFlow · TF Recommenders · OpenCV · LangChain · CrewAI
Tools & Databases
Git · Docker · Jupyter Notebook · Unix · PostgreSQL · REST APIs
Platforms
AWS · Docker
Education
Master of Science in Information Technology and Management
The University of Texas at Austin
2026-2027
Bachelor of Science in Physics
Texas A&M University
2021-2025
Astronomy concentration with minors in Computer Science and Mathematics
Experience
- Updated the TARDIS environment by refactoring pandas/NumPy across the codebase for compatibility with pandas 3.0, allowing the team to adopt newer package versions without breaking existing functionality
- Identified dead code in the plasma module via pytest coverage reports to consolidate findings into architecture documentation, improving maintainability for future contributors
- Closed test coverage gaps by writing pytest unit tests with numpy.testing and pandas.testing assertions against regression data, strengthening validation of plasma properties previously covered only by integration tests
Google Summer of Code open source contributor with the TARDIS RT collaboration, working on the plasma module of TARDIS — open-science software for simulating and analyzing supernovae with Monte Carlo radiative transfer.
Key Accomplishments
- Optimized a database of 200+ physics demos using SQL and Python
- Integrated the demo ordering system with the Google Calendar API to notify employees of demo delivery details
- Built a CrewAI multi-agent system using LangChain and RAG algorithms to parse Google Docs and suggest demos with continuous updates
- Developed Python workflows for automated text summarization and resource classification
The Physics Lab Center (PLC) is responsible for designing, building, and maintaining physics demonstrations used in lectures and outreach events. Throughout my time at the PLC I've maintained our storeroom and workshop which houses over 200 demonstrations, covering topics such as thermodynamics, electricity and magnetism, classical mechanics, and modern physics. Another aspect of the job was updating the website where professors order demonstrations with demo descriptions and pictures. As I continued to work at the PLC I noticed many areas where updated technology could optimize a lot of our work on the employee side along with maximizing ease of use on the professor's side. This is why I’m currently working on updating the technology stack of our website and incorporating predictive models and agentic AI to better serve relevant demonstrations to professors. I’ve written more about these projects here (website update) and here (PLC agentic AI). Properly documenting the storage and use of these physics demonstrations is especially important because these demonstrations are also presented during high-profile events like the TAMU Physics and Engineering Festival, which attract over 5,000 attendees annually.
Key Accomplishments
- Developed autonomous rover navigation software in C++ and Python with OpenCV following the V-model for iterative testing
- Designed and updated Requirement Verification Matrices and subsystem documentation
- Managed a $1,000 subsystem budget and coordinated hardware procurement and test planning
At AggieSat Laboratory, I collaborated with a 30-member team to develop a project proposal for NASA’s PRISM solicitation. We prototyped a network of devices to establish a "lunar GPS system" capable of supporting future missions focused on lunar navigation, regolith composition studies, and building a lunar base camp. Additionally, I served as the Command and Data Handling (CDH) Subsystem Lead for the 20-member GeoRGE team in an internal rover competition. In this role, I led the design of the CDH subsystem by completing ranking verification matrices, budget sheets, and test verification documents. My team designed and programmed autonomous rover software utilizing computer vision algorithms and stereoscopic vision, which enabled the rover to navigate terrain modeled after the surfaces of Mars and the Moon. Finally I joined the Machine Learning team where my team and I researched how to use ML models in space engineering systems and started ideating on models that improve the efficiency of the MUSIC algorithm - which is a frequency estimation and radio direction finding algorithm.
Key Accomplishments
- Analyzed 1,000+ JWST observations using Python (NumPy, Pandas, SciPy)
- Applied MCMC fitting with emcee and Astropy to model stellar populations
- Containerized pipeline with Docker, reducing setup from 8 hours to 1 hour
AggieSTAAR was a 10-week undergraduate research program where I did galaxy evolution research under Dr. Casey Papovich. The type of astrophysical research I did was essentially data engineering and data science — I cleaned large datasets, performed exploratory data analysis work on James Webb Space Telescope galaxy data, and I fit the galaxies relevant to my research to models of galaxy evolution. I also got to hone my data visualization skills by presenting my research at the 245th American Astronomical Society Meeting. I go more into the details of my research here.
Key Accomplishments
Projects
The projects more tied to my work experience are documented soley in the projects page. This is where I want to highlight some personal and student org projects.
*project pages and WIP blogs for all projects will be coming soon!
Album of the Week is a book club, but for albums! I joined the org in its infancy and was one of the first automation committee members. On this committee I worked with a team of 3–5 people to develop a codebase which fully centralized the voting, nomination, rating, and review process on our organization's website. Previously the nomination to rating process was done through Google forms that had to be sent out to members mutiple times throughout the week every week. On top of that, our organization facilitates a lot of music discussion on our Discord server, so the automation team also developed a Discord Bot that interfaced with the backend of our Website.
Key Accomplishments
- Developed JavaScript Discord bot with PostgreSQL integration for tracking posted song recommendations and updating weekly Spotify playlist
- Contributed to full-stack React.js + Express.js web app supporting album nominations, voting, and reviews
- Implemented asynchronous event handling and fail-safe scripts, applying AGILE sprints for iterative feature deployment
Album-Per-Year (not the final name) is a web app that lets users connect their Spotify or Last.fm account, enter their birthday, and then select one of their favorite albums released for each year they’ve been alive. At the end, the app generates a album chart. Read more about the project and where it currently is at here.
Key Accomplishments
- Built full-stack web app using React.js, Express.js, and FastAPI for users to generate shareable album charts
- Created PostgreSQL database of 40,000+ ListenBrainz profiles for large-scale recommendations
- Trained a Two Tower Recommendation model with TensorFlow Recommenders to serve relevant albums to users; trained and deployed via AWS SageMaker
- Implemented OAuth login for Spotify and Last.fm and generated real-time feature vectors for model inference
Cricanalytics is a streamlit dashboard that visualizes sabermetrics style analysis applied to cricket games from the IPL (Indian Premier League) during the 2023-2024 season. This was a project hosted by the Aggie Data Science club that really interested me because Cricket is a very important sport in my family and my community, but at the time I only understood the base rules of the sport. Diving deep into analysis of Cricket while learning more about the official data science industry terms and visualization processes was a really interesting way to learn about the sport and connect more with my Cricket-head dad. When I orginally worked on this project in 2024 the PM scraped the data we initially used, and I'd like to come back to this project with 2025 season data as a way to practice ETL process on data I've scraped myself.
Key Accomplishments
- Built an interactive Streamlit dashboard to analyze IPL 2023–2024 player performance
- Performed EDA with Pandas and created visualizations with Seaborn
- Developed a custom expected wickets metric using K-means clustering to evaluate bowling performance with data-driven insights