Work
See ResumeChipper Cash
Software Engineer II (2023-Present Full-Time • 2020-2022 Intern, Part-Time)
I have worked at Chipper full-time since August 2023 and in several internships. At Chipper I have worked with a focus on web and mobile platforms, ensuring users have a reliable experience with the application, and that mobile developers (including myself) build features with the right tools and patterns.
Some of the work I have done includes:
In my current role, I have implemented a visual refresh for our Virtual Cards feature in our mobile application and also introduced support for allowing users to set recurring payments across different product areas. The Virtual Cards product is one of our most used products.
I also lead initiatives for improving the reliability of our mobile platform, including implementing an error-tracking and logging service, and optimizing client fetching and caching strategies to make the app feel fast, and to decrease server load across users (10M).
I am also actively working with backend engineers to extend our payment functionality to the web - beyond just mobile.
More from 2020-2022
In 2021 Fall, and 2022 Spring I worked on CLI tools and functionality around Redux Toolkit Query HTTP API client to allow frontend teams to automatically generate APIs for interacting with our backend.
Furthermore, I designed and built a system on top of RTK Query for synchronizing application state with long-running (or asynchronous) operations using polling strategies.
In 2020 Fall, I built a declarative deep-linking system to allow mapping external URLs to specific screens and custom actions within the Chipper React Native mobile application.
This project allowed for growth and marketing teams to perform targeted campaigns with continuity directly within the app.
The system has since been widely adopted for deep-linking from push notifications, from our user messaging center, from emails to power frictionless onboarding, in-app banners and payment link features (for example, integration to allow users to receive payments from X Tips).
In 2020 Summer, for my first internship project, I built a web app allowing users to view receipts across different products and services offered on the platform such as P2P payments, bill payments and more.
This was a full-stack project that involved building re-usable receipt component containers, and then custom data fetching and transformation logic for different receipt types.
BlackRock
Software Engineering Intern (2022 • 2021)
At BlackRock, I participated in two Software Engineering internships, first in 2021 and then in 2022.
In 2022 Summer, I worked with a team to extend a multi-client accounting web platform to support the creation and configuration of accounting workspaces for legal entities.
This was a full-stack project where we used Angular, RxJS and AGGrid on the frontend to implement new panes and tables on the accounting portal to allow users to create and configure legal entities.
On the backend, we used the Spring Boot Java framework, adding new Data Access Object (DAO) methods and REST endpoints for storing and retrieving legal entities and their configurations.
I collaborated with fellow interns, my manager and other software engineers and UI designers to plan and build the new features on the platform, delivering several live demos to stakeholders and sponsors and a final presentation across the wider engineering team.
More from 2021
In 2021 Summer, I worked on a team to implement a Sequence Clustering Machine Learning algorithm in Python to identify patterns/clusters across user sessions on web applications. The algorithm was determined and built by reading research papers to understand the underlying concepts of 1st-order Markov chains and Expectation Maximization.
I proposed and implemented the API design for the algorithm with inspiration from SciKit Learn's models for a familiar developer experience. I also organized and packaged the algorithm as a Python package to allow for easy installation and use across different projects.
To allow for experimentation and collaboration on different hyperparameter initialization strategies, I designed and built a plugin-like system to allow teammates to simultaneously develop different initialization modules. Finally, I also built support for inference and basic serialization and deserialization of trained models to allow the loading of pre-trained models for inference and visualization.
A significant personal achievement was optimizing the algorithm for performance, cutting down execution time by 91% while maintaining clustering quality. This was achieved through several rounds of NumPy vectorization and matrix transformations to remove unnecessary loop iterations compared to the naive approach.
Finally, I implemented a Jupyter Notebook to showcase a step-by-step workflow/guide to loading training data, fitting the model, and visualizing the clusters with different visualizations such as: heat maps, Sankey diagrams, step diagrams. The visualizations were implemented together with my teammates using seaborn.