SFI Foundation 2026–27
CSP 2026–27 capstone continuing the SFI Foundation prototype with searchable motorsports safety standards, ML-assisted discovery, equipment detection, personal gear tracking, and staff tools.
Student capstone · In development. This project explores a clearer way to discover and organize motorsports safety information. It is not an official SFI Foundation product and does not replace official SFI standards, labels, or PDF documents.
A continuation of our motorsports safety modernization prototype, focused on helping people search standards, understand likely matches, organize equipment, and revisit important certification information.
Experience: discover → understand → inspect → save → revisit
Project focus: searchable standards, assisted discovery, equipment recognition, personal gear tracking, and staff-oriented workflows.
Core experience
Four connected parts of the user journey, from finding a standard to returning to saved gear later.
Browse categories or search specification records in plain language instead of relying only on exact specification numbers.
A TF-IDF + LinearSVC classifier suggests likely specification matches from a free-text equipment description.
Browser-side TensorFlow.js models explore image and camera-based equipment recognition as an assistive discovery tool.
Users organize personal gear, revisit certification information, and ask the site chatbot questions about the available specification data.
From problem to product direction
Users can face dense lists, unfamiliar specification numbers, and multiple documents when determining which safety standard applies to a piece of equipment.
Bring structured specification data, plain-language search, ML suggestions, gear tracking, and guided tools into one consistent frontend backed by a Flask API.
System flow
A simple view of how the browser, API, data layer, and assisted-discovery tools work together.
Jekyll and JavaScript provide search, equipment detection, My Gear, authentication views, and chatbot interactions.
Backend routes handle authentication, SFI specification endpoints, classifier requests, chatbot requests, and gear operations.
SQLAlchemy and SQLite organize structured specification records and user-linked prototype data during development.
LinearSVC matching, TensorFlow.js detection, and Gemini-assisted questions help users narrow down relevant information.
The goal is a single workflow where the browser experience and backend services can evolve together instead of feeling like separate demos.
Technical foundation
Jekyll + JavaScript for static content and interactive client-side features.
Python Flask APIs for authentication, specifications, chatbot requests, and gear operations.
TF-IDF + LinearSVC text classification and TensorFlow.js experiments for assisted equipment discovery.
SQLAlchemy persistence with SQLite in development, plus a Gemini-backed chatbot using compact specification context from the backend.
Team
Two Scrum Masters guide project coordination while four technologists/developers build and refine the product experience.
Scrum Masters
Scrum Master
Scrum Master
Technologists / Developers
Technologist / Developer
Technologist / Developer
Technologist / Developer
Technologist / Developer