Logan

Logan started as an English teacher for topics I actually care about. After each lesson it summarizes what happened, draws conclusions, and changes the next exercises based on what worked or bored me.

Language learning is often boring and disconnected from real interests. Motivation drops when lessons do not adapt to the learner.

If a teacher remembers previous lessons and reacts to how someone learns, language learning can become less mechanical.

• Lessons based on goals, interests, and user context.
• Reflection loop after each lesson: summary, conclusions, and changed future sessions.
• Vocabulary, idioms, quizzes, and homework.
• Integrations with Make, Airtable, Shortcuts, ElevenLabs, and a path toward Telegram/Siri.

LangChain/GPT runs lessons and reflection after a session. Memory lets Logan change explanations and exercises based on what worked before.

Python, LangChain, GPT, Firebase, Airtable, Make, Shortcuts, ElevenLabs, and Flask/API as execution layer.

• Private version tuned for real usage before opening it to users.
• Reflection as a mechanism for teaching the teacher, not only the student.
• System integrations on Mac/iPhone so learning can happen in daily context.

Working private system with lessons, quizzes, homework, post-session reflection, and a plan for a user-facing version.

Finish onboarding, session management, long-term memory, and an external user version.