[cite_start]Based on the document provided, the title is "**AI-based drop-out prediction and counseling system**"[cite: 3].

[cite_start]UniGrit is a project participating in the Smart India Hackathon 2025 that addresses the problem of student dropouts[cite: 1, 2, 3, 4]. [cite_start]The project, with Team ID UniGrit, falls under the "Smart Automation" theme and is a software solution[cite: 5, 6, 8].

### **Proposed Solution and Technical Approach**

[cite_start]UniGrit's solution is a consolidated digital dashboard that uses an AI-based system to predict student dropouts and provide counseling[cite: 4, 24]. [cite_start]The system automatically takes in student data, applies rule-based thresholds to identify at-risk students, and sends notifications to mentors and guardians for early intervention[cite: 24].

Key features of the system include:
* [cite_start]An **Agentic Counseling AI** that generates personalized intervention strategies and communication drafts based on identified risk factors[cite: 25].
* [cite_start]An **Intuitive Educator Dashboard** that centralizes student risk data, trends, and AI-suggested actions for review by educators[cite: 26].
* [cite_start]**Proactive Alerts** that send timely and configurable notifications to mentors and guardians[cite: 27].
* [cite_start]**Empowered Human Judgment** that allows educators to easily review and approve the AI-generated plans[cite: 28].

The technical approach involves a backend and a frontend. [cite_start]The **backend** includes a centralized data warehouse that ingests data like student profiles, test scores, fees, and attendance[cite: 44, 50, 51, 53, 56]. [cite_start]An analytics engine within the backend uses a predictive model, rule checks, and anomaly and trend analysis to flag high-risk students[cite: 47, 49, 59, 61, 62, 69]. [cite_start]A proactive intervention agent then triggers actions and sends alerts through a notification scheduler[cite: 71, 73, 75]. [cite_start]The **frontend** consists of a mentor dashboard, student portal, and a system for guardian notifications[cite: 83, 86, 88]. [cite_start]The system is built using technologies such as MongoDB, Express.js, Node.js, and scikit-learn[cite: 91, 94, 95, 98, 99].

### **Feasibility, Viability, and Business Potential**

[cite_start]The UniGrit system is designed to be feasible and viable[cite: 104, 109]. [cite_start]It's described as "plug-and-play" and cost-efficient, as it uses existing institutional data and requires minimal training[cite: 105, 107, 110]. [cite_start]The model is proven, with similar apps like Civitas, Starfish, and EWS showing strong adoption[cite: 114, 121].

[cite_start]The business potential for UniGrit is a **Low-Cost SaaS (Software as a Service) Model** offered as a cloud subscription to institutions[cite: 130]. [cite_start]The project targets partnerships with public universities and schools under government schemes, as well as collaborations with NGOs in the education sector[cite: 131, 132, 133].

---
### **Impact and Benefits**

[cite_start]The UniGrit project is intended to have a significant impact and provide ben

1. SMART INDIA HACKATHON 2025
•Problem Statement ID –25102
•Problem Statement- Title-AI-based
drop-out prediction and counseling
system
•Theme- Smart Automation
•PS Category- Software
•Team ID-
•Team Name- UniGrit

2. UniGrit
2@SIH Idea submission- Template
UniGrit
Automated Intervention Planning Personalized Resource Curation
Autonomous Student Notifications
Continuous Model Retraining
Mentor-Guardian Meeting Scheduling
Conversational Support & Q&A
Other Features
•Agentic Counseling AI: Generates personalized intervention strategies and draft
communications based on risk factors.
•Intuitive Educator Dashboard: Consolidates student risk, trends, and AI-suggested actions
for review.
•Proactive Alerts: Dispatches timely, configurable notifications to mentors and guardians.
•Empowered Human Judgment: Facilitates easy review and approval of AI plans,
empowering educators.
Proposed Solution
A consolidated digital dashboard that automatically ingests student data, applies rule-based thresholds to identify at-risk
students, and dispatches regular notifications to mentors and guardians for early, data-driven intervention.

3. TECHNICAL APPROACH
3@SIH Idea submission- Template
UniGrit

4. FEASIBILITY AND VIABILITY
4@SIH Idea submission- Template
UniGrit
Feasibility Business Potential
Viability
•Plug-and-Play
•AI-assisted
•Cost efficient
•Agentic AI PoweredConversational assistant + Mentor style guidance
•Proven Model
•Community Validation
•Engagement-first
•Students
•Teachers/Admins
•Institutions
•Low- Cost SaaS Model

•Institutional Partnerships
•CSR & NGO Collabs
Low adoption by
teachers
Data scattered
across files
Delayed
intervention
We included Risk vs Solution to show we’ve anticipated real
challenges and already designed ways to overcome them
Agentic AI answers
queries
AI alerts & guides
before dropout.
Automated Excel
unified dashboard.
Ask AI: “What are my chances of dropout this semester?” → get
personalized advice.
Ask AI: “Which students are near dropout criteria?” → get
real-time filtered list.
Use Cases
Retrieval of data from seperate institutional datasets.
Risk prediction using ML + rule-based early alerts
Uses existing data (no new infra)+minimal training needed.
Similar Apps (Civitas,Starfish,EWS) show strong Uptake
Feedback Loop -> Learning Mechanism-> The Result
Student chatbot mentor + teacher AI advisor drive
continuous interaction.
Preventive measures reduce dropout rates, improve academic
performance.
Offer as a cloud subscription for Institutes
Target public universities, schools under
Govt. schemes
NGOs working in Education(Pratham) can
adopt it to support rural students
RISK VS SOLUTION

5. IMPACT AND BENEFITS
5@SIH Idea submission- Template
UniGrit
1

Participation
•Empowers educators, mentors & guardians to
act early
•Builds shared accountability in preventing
dropouts
2
Governance
•Unified dashboard across
attendance, scores & fees
•Enables data-driven routing &
institutional decisions
3 Trust
•Accurate, tamper-proof academic &
attendance records
•Boosts compliance, transparency &
stakeholder confidence
1
Speed
• Flags risks weeks before results
•Enables timely & precise interventions
2
Savings
•Cuts dropout-related costs & re-enrollment
losses
•Optimizes institutional & family resources
3
Experience
•Simple, visual dashboard with risk indicators
•Boosts mentor–student engagement & trust
4
Growth
•Scalable, adaptable across institutions
•Establishes leadership in AI-driven education
4
IMPACTS
BENEFITS
Inclusion
•Early detection of vulnerable learners across
backgrounds
•Ensures equitable support for every student

6. Agentic AI
RESEARCH AND REFERENCES
@SIH Idea submission- Template 6
UniGrit
Feature Our SystemStarfish(EAB)Civitas Learning
Administrator
Dashboard
Real Time risk
detection
Intervention
Pathways
Chatbot
(24/7 Support)
AI-powered
Counselling
Proactive Alerts
Perfromance
Monitoring
UI/UX for the
application
QR to the
application