AI-Predictive Vigilance Tool

Sentri is a specialized machine-learning platform for the Poway Recovery Center. By analyzing 14-day data vectors of sleep, stress, and mood, Sentri foresees relapse windows before they occur and automates intervention strategies to protect long-term sobriety.

Sentri Logo
ML Neural Network Active

Random Forest Classifier v1.0

SENTRI

Predictive Sobriety Guard

◈Relapse Risk Model: ML algorithm analyzing historical logs to generate a real-time risk percentage.
◈Mood-Biometric Dashboard: High-speed entry for tracking sleep quality, stress levels, and craving intensity.
◈Automated Alert System: Triggers immediate "Intervention Protocol" notifications when risk thresholds are met.
◈Dynamic Roadmap: Milestone visualization that adjusts "Difficulty Ratings" based on your personal data.
◈NLP Sentiment Triage: Analyzes raw text input to identify and route medical emergencies instantly.
CSP Architecture

Sentri utilizes a Recurrent Neural Network (RNN) architecture to process a List of user inputs.

The Procedure iterates through the user's historical data, using Selection logic to determine which intervention resource from the database best matches the current risk level.

Vigilance Impact
Eliminates the "Blind Spot" in recovery by providing a functional diagnostic tool that warns users 48-72 hours before a potential relapse occurs.


Project Overview

Sentri is an AI-powered sobriety tracker designed to solve the “blind spot” problem in addiction recovery. Traditional tracking tools are reactive; they count days after they have already passed, leaving users vulnerable during high-risk periods they can’t see coming.

Sentri changes that by using machine learning models to analyze a user’s biometric data, sleep patterns, and mood logs. By calculating a “Relapse Risk Score” based on historical patterns, the platform proactively identifies potential crises and pushes personalized intervention strategies—ranging from suggested meetings to immediate support resources—before the user reaches a breaking point.

Key Objectives

  • Predictive Awareness: Move beyond static day-counting to provide real-time alerts when the system detects a downward trend in wellness.
  • Proactive Intervention: Use NLP and behavioral modeling to offer specific, actionable support (e.g., “Call your sponsor” or “Try a breathing exercise”) exactly when it is needed.
  • Personalized Roadmap: Replace generic milestones with a dynamic assessment tool that adjusts the difficulty of the recovery journey based on the user’s specific historical struggle points.
  • High-Fidelity Triage: Implement an automated sentiment analysis system to prioritize critical emergencies, ensuring that users in crisis are routed to immediate help rather than standard resources.

By moving away from static checklists and toward active, AI-assisted monitoring, Sentri provides a digital safety net that stands guard over the user’s sobriety 24/7.