Chronos

• 

  Daily briefing modal showing the Principal’s plan adjustments and reasons.

• 

  Dashboard overview with key learning metrics and quick actions.

• 

  Plan view with life interruptions, adaptive timeline, and mode selection driven by the Principal.

## Inspiration Study plans often fail when real life changes. Late meetings, poor sleep, or stress can derail the best schedule. I wanted a learning system that adapts to life events instead of blaming the learner.

## What I learned

• A good learning system needs a decision layer, not just a task list.
• User “complaints” are valuable signals of life pressure.
• Recommendations must be explainable to build trust.

## How I built it Chronos has three layers:

1. Life Event Capture: natural language complaints are stored as structured life events.
2. Principal Decision Engine: combines life events, learning records, and weekly plans to adjust today’s mode and tasks.
3. Daily Briefing: shows the adjusted plan with reasons so users can accept or modify.

We optimize for plan–reality alignment: [ \text{Plan}^*=\arg\min_P \left(\alpha \cdot \text{LifePressure}(t) + \beta \cdot \text{PlanDeviation} (P,R)\right) ]

## Challenges

• Converting free‑form complaints into structured signals.
• Making adjustments explainable rather than a black box.
• Keeping real‑time updates lightweight for users.

Built With

• aiosqlite
• flet
• google-gemini-api
• python
• pyvis
• sqlite