Nicholas Gardella

A post-analysis app for Taguette. See video for instructions. I built this to use for qualitative research.

A playground for experimenting with ways to improve the experience of highlighting selections of HTML as in Taguette. This is an alternative to my LoQDA approach, which abandons trying to coerce every format to HTML, opting for PDF instead.

The seed for my unrealized dream of a free local-first web-based Qualitative Data Analysis app centered around the PDF. It is heavily inspired by Taguette and OpenQDA. For now, its major capability is to create and save hightlight annotations directly into a PDF file. I don't have time to work on this at the moment but really wish I did. PDFs are harder than you'd think!

A web-based Bluetooth Low Energy central device driver for the Heart Rate Service. It can collect heart rate and R-to-R wave intervals from any generic sensor, and can collect Pulse-to-Pulse intervals from the Polar OH1 through a driver I reverse-engineered from Polar's Kotlin & Swift SDKs. It can collect and save data to a local SQLite database from multiple sensors simultaneously.

My first attempt at a web-based Bluetooth Low Energy central device driver for the Heart Rate Service. See this video for a demonstration. It can collect heart rate and R-to-R wave intervals generic sensors, but is designed for the Polar H10, from which it can collect and display electrocardiography (ECG) at 160Hz. I also reverse-engineered this driver from Polar's Kotlin & Swift SDKs.

A web-based Bluetooth Low Energy central device driver for the Generic Health Sensor Service. Currently it displays Galvanic Skin Response data from my prototypes for the Gardella GSR-To-Go peripheral device and saves battery data so I can calibrate the battery percentage estimator. I also tried out TinyBase for this project.

A prototype for counting steps using the accelerometer in an Android device. The application tries to assess when a step is taken using an exponentially weighted moving average of the 3D acceleration vector.

An embeddable, accessible form control for entering the NASA's Task Load Index scale items. I designed this for easy embedding in Qualtrics and other survey platforms to improve the consistency of delivering the NASA TLX scale. It is compatible with multitouch, screen readers, keyboard controls, and mouse inputs. I created this to aid a peer in their research on the NASA TLX. They used my code as a basis to build software for their research studies. This helped me improve my D3.js, SVG, and accessibility skills!

An SVG-based interactive visualizer for scoliosis curvature. I made this rough prototype as a proof-of-concept for orthopedic surgeons who wanted a way to capture patient expectations about how their spines would look after surgery. This helped me learn a lot about SVGs!

A block-scheduling planner inspired by Cal Newport's Time Block Planner. The data can be saved or loaded from the browser, and the blocks support context menus, drag-and-drop, splitting, deletion, labeling, and color choice. I used this project to learn React, Redux, drag and drop, Material UI, and many other web technologies.

A proof-of-concept app I built to bring together Transformers.js and libSQL WASM. It vectorizes sample data and performs RAG inside SQLite, all locally inside the user's browser. I plan to use this approach in the future to add local AI features to my front-end CRUD apps.

A Svelte app to navigate my grades for my college and graduate courses. I used this project to teach myself Svelte!

My version of a Vanilla JS movie app I taught interns to build at Salesforce in 2022.

A typescript utility that scrapes public endurance workouts from 8020 Endurance's Workout Library, turns them into structured text workouts, and uploads them to your account on Intervals.icu, an endurance training planning app and TrainingPeaks competitor. I built this because I owned Matt Fitzgerald's books and associated training plans but wanted to get them into my own training software easily, without having to write out nearly 500 workouts manually. I shared this on the forum for Intervals.icu and other users have found it helpful as well.

A Cloudflare Workers implementation of HenrikJoreteg's github-secret-keeper with added rate limiting. This allows for a serverless approach to authenticating static apps using GitHub. I built this to enable me to build multi-user applications with limited hosting costs.

A native windows app written in Go to bypass CORS restrictions using a local proxy made for users without Node.js or other command-line experience (unlike the original UnCORS) Download from the GitHub releases

I built this app in Swift and Swift UI in 2020. it is not published in the App Store, but Strong is probably better anyway! Sadly the code is held hostage in my VT GitHub account, which I can no longer access 😔