Family
Thompson Sampling
in-developmentApache 2.0rlretail-facing
Budget Guardian.
RL agent for budget-anomaly alerts. Thompson Sampling over user spending patterns. Continuous online learning.
Foundation · Thompson Sampling
What it does
The honest description.
RL agent for budget-anomaly alerts. Thompson Sampling over user spending patterns. Continuous online learning.
Architecture
How it works.
User monthly budget by category, spending-to-date per category, budget-utilization %, 6-month category spending history, income baseline. Budget status alert (on-track | warning | exceeded) + recommended action (adjust budget | increase saving | review category spending) + confidence score.
Training data & drift
What it learned from, and how we watch it.
Source
Inputs: User monthly budget by category, spending-to-date per category, budget-utilization %, 6-month category spending history, income baseline.
Drift baseline · current quarter
Drift baseline not yet documented for this model.
What it's NOT
- Thompson Sampling RL agent.
- Read-only monitoring (no budget mutations).
- Online learning from user acceptance/rejection of alerts — no batch training.
Who calls this model
Across the house and the API.
Internal routes
No public routes documented yet.
Featured in
/features/spending-budgetsVersion history · immutable once published
Every release, traceable.
Data Mukodi holds about you is subject to DPDP §12 erasure (use /consent in the app), and any concern about this model can be raised via [email protected] under DPDP §13 grievance, or via /security.