The Background: I built SkyLaunch to solve the systemic inefficiencies I faced as a Part 61 flight student. While flexible training is great, it suffers from a total lack of data structure compared to Part 141 schools.
The Problem:
State Management: Instructor turnover is high. When I switched CFIs, my "training state" (proficiency in specific maneuvers) relied on paper logbooks or memory, leading to wasted hours re-demonstrating skills I had already mastered.
The "Black Box" of Costs: I was flying blind financially, with no way to forecast "Time-to-License" based on my actual learning velocity.
The Solution: SkyLaunch is an open-source flight assistant that acts as a progression engine. It maps your flight maneuvers to specific FAA Airman Certification Standards (ACS), visualizing your proficiency in real-time. It effectively decouples your training data from your instructor.
The Tech:
Frontend: React Native (optimized for iPad Mini).
Backend: Supabase (Postgres RLS for student privacy + Realtime for Student-CFI sync).
Logic: A custom scheduling engine that checks weather minimums (METARs) against your personal proficiency to issue "Go/No-Go" recommendations before you drive to the airport.
Feedback Request: I’m looking for feedback on the scheduler logic (lib/schedulerLogic.ts)—specifically how to better model "skill decay" for maneuvers that haven't been practiced in X weeks.
Links:
Repo: https://github.com/unshah/SkyLaunch.io
iOS: https://apps.apple.com/us/app/skylaunch-flight-training-app/...