WellNest: A Django + MongoDB Powered Family-Centric Wellness Platform

Pippalla Sri Kiran — Mon, 27 Apr 2026 13:31:46 +0000

WellNest is a Django-first full-stack wellness platform powered by MongoDB for real-time, user-centric data management. It integrates daily wellness tracking, AI-driven emotional insights, and family-supported interactions into a unified system. Django serves as the core orchestration layer, providing REST and GraphQL APIs for seamless functionality. MongoDB enables flexible storage of dynamic data such as check-ins, goals, messages, and insights. The platform uses lightweight AI for sentiment analysis, risk detection, and personalized recommendations. Overall, WellNest demonstrates a family-centric, preventive approach to digital wellness using modern full-stack technologies.

Team Members

This project was developed by:

@ambareesha_vishweshakuma — A.Vishwesha Kumar
@ashfyx716 — Ashfaq Sayeed Shaik
@pippalla_srikiran — P.Sri Kiran

We would like to express our sincere gratitude to @chanda_rajkumar for their valuable guidance and support throughout this project. Their insights into system design, architecture, and development played a crucial role in shaping WellNest.

The Problem We Set Out to Solve

Many middle-aged women silently experience declining physical and mental well-being due to continuous family responsibilities, stress, and lack of self-care. In most households, early signs like poor sleep, fatigue, or emotional strain go unnoticed because there is no simple way to track or support them daily.

We built WellNest to solve this. It enables mothers to log daily wellness signals and receive instant AI-driven insights, while allowing family members to actively support them through simple, meaningful interactions—creating a continuous, real-time wellness support system.

But the real strength of WellNest is not just its AI capabilities — it lies in how MongoDB powers the entire system, from storing daily wellness data to enabling real-time insights and meaningful family interactions.

The WellNest Approach

WellNest is designed as a family-centric wellness platform that transforms health tracking from an individual task into a shared responsibility. Instead of relying solely on self-discipline, the system combines daily wellness tracking with AI-driven insights and family-supported interactions.

At its core, WellNest allows users to log simple daily signals such as mood, sleep, and activity. These inputs are analyzed using lightweight AI to detect patterns, assess risk levels, and generate personalized suggestions. At the same time, family members are included in the loop, enabling them to provide encouragement, support, and timely interventions.

This approach shifts wellness from a reactive process to a preventive and collaborative system, where early signals are identified and addressed before they escalate.

Why Django + MongoDB Was the Right Choice

WellNest uses Django and MongoDB to balance strong backend orchestration with flexible, real-time data handling. Django manages APIs, authentication, business logic, and AI workflows within a single, scalable backend, ensuring the system remains centralized and maintainable. MongoDB fits naturally with the platform’s dynamic data, including check-ins, insights, goals, and family interactions, using a flexible document-based model. This allows the schema to evolve as features grow without rigid constraints. It also enables real-time analytics through aggregation pipelines, ensuring responsive dashboards. Together, Django powers the system backbone while MongoDB drives the data and intelligence layer, making this stack ideal for WellNest’s family-centric wellness model.

System Architecture Overview

WellNest processes user inputs through Django, applies AI analysis, and stores data in MongoDB for real-time insights.
This architecture enables a smooth flow from data capture to personalized recommendations and family support.

Data Model in MongoDB

users – stores user profiles and preferences
checkins – daily wellness entries (mood, sleep, activity)
goals – user-defined wellness goals and progress
care_pulses – family support messages and interactions
ai_insights – generated insights, risk scores, and suggestions
predictions – AI outputs and classification results
family_links – relationships between users and family members

from functools import lru_cache
from pymongo import MongoClient

MONGO_URI = "mongodb://localhost:27017/"
MONGO_DB_NAME = "wellnest"

@lru_cache(maxsize=1)
def mongo_client():
    client = MongoClient(
        MONGO_URI,
        serverSelectionTimeoutMS=5000,
        connectTimeoutMS=5000,
        socketTimeoutMS=10000,
        retryWrites=True,
    )
    client.admin.command("ping")  # verifies connection
    return client

def mongo_db():
    return mongo_client()[MONGO_DB_NAME]

The 7 Aggregation Pipelines That Power Insights

Daily Check-in Trend – Processes check-in data to show mood, sleep, and activity patterns over time on the dashboard.
Risk Score Distribution – Analyzes AI insights to calculate how many users fall into low, medium, or high-risk categories.
User Activity Tracking – Measures how frequently users log check-ins and interact with the platform.
Family Interaction Metrics – Tracks care_pulses and support messages exchanged between family members.
Goal Completion Rate – Evaluates how many wellness goals are completed versus pending.
AI Insight Trends – Identifies recurring emotional or behavioral patterns from AI-generated insights.
Alert Summary Pipeline – Highlights recent high-risk cases that may require attention or support.

AI & ML Integration in WellNest

**
WellNest uses a set of lightweight AI and machine learning components to generate meaningful wellness insights without being complex or clinical. When a user submits a daily check-in, the system analyzes inputs such as mood, sleep, and text reflections to understand emotional patterns and overall well-being.

The platform combines multiple components for this process. The TextInsightEngine extracts emotional signals and generates supportive responses from user text. The ml_mood_signal model evaluates check-in patterns to compute a wellness risk score and classify it as LOW, MEDIUM, or HIGH. Based on this context, build_ai_gentle_goal creates personalized and achievable wellness goals.

Additionally, an optional Nesti LLM layer enhances chatbot interactions by generating richer, conversational responses when enabled, while fallback logic ensures reliability at all times.

These AI and ML components work together to provide context-aware insights, personalized recommendations, and continuous emotional support, making WellNest a practical and preventive wellness platform.

# --- Text Insight Model: TextInsightEngine ---
from textblob import TextBlob
from collections import Counter

class TextInsightEngine:
    def analyze(self, text: str) -> dict:
        blob = TextBlob(text)
        polarity = float(blob.sentiment.polarity)
        subjectivity = float(blob.sentiment.subjectivity)

        if polarity > 0.2:
            mood = "positive"
        elif polarity < -0.2:
            mood = "negative"
        else:
            mood = "neutral"

        keywords = [str(w).lower() for w, _ in Counter(blob.noun_phrases).most_common(5)]
        return {
            "mood": mood,
            "polarity": polarity,
            "subjectivity": subjectivity,
            "keywords": keywords,
            "recommendation": self._recommend(mood),
        }


# --- Risk Model: ml_mood_signal ---
def ml_mood_signal(mother_id: str, latest: dict | None = None) -> dict:
    rows = list(_db()["checkins"].find({"user_id": mother_id}).sort([("entryDate", -1)]).limit(7))

    if latest and not rows:
        rows = [latest]

    stress_like = sum(1 for r in rows if (r.get("mood") or "").upper() in {"STRESSED", "SAD"})
    tired_like = sum(1 for r in rows if (r.get("mood") or "").upper() == "TIRED")
    low_sleep = sum(1 for r in rows if (r.get("sleepQuality") or "").upper() == "POOR")
    conflict = sum(1 for r in rows if bool(r.get("bertConflictsWithManual", False)))

    n = max(len(rows), 1)
    risk_score = (
        (stress_like / n) * 0.55 +
        (tired_like / n) * 0.20 +
        (low_sleep / n) * 0.20 +
        (conflict / n) * 0.05
    )

    risk_score = max(0.0, min(1.0, round(risk_score, 2)))

    if risk_score >= 0.65:
        risk_level = "HIGH"
    elif risk_score >= 0.35:
        risk_level = "MEDIUM"
    else:
        risk_level = "LOW"

    return {"riskScore": risk_score, "riskLevel": risk_level}


# --- Goal Generation Model: build_ai_gentle_goal ---
def build_ai_gentle_goal(
    mother: dict | None,
    latest: dict | None,
    hint: str | None = None,
    ml: dict | None = None
) -> str:
    mood = ((latest or {}).get("mood") or "NEUTRAL").upper()
    ml = ml or {}
    risk_level = str(ml.get("riskLevel") or "LOW").upper()

    fallback_by_mood = {
        "STRESSED": "Ask mom to do a 5-minute breathing pause and drink a glass of water.",
        "SAD": "Send mom a warm note and suggest a short walk near sunlight.",
        "TIRED": "Encourage mom to rest 20 minutes and hydrate before the next activity.",
        "HAPPY": "Invite mom to repeat one healthy habit that made today feel good.",
        "CALM": "Help mom protect her calm with a light stretch and quiet tea break.",
    }

    if risk_level == "HIGH":
        return "Guide mom through 4-4-6 breathing and ask for one immediate support need."

    return fallback_by_mood.get(
        mood,
        "Ask mom to do one gentle self-care step and check in tonight."
    )


# --- Optional LLM Configuration (Nesti Layer) ---
import os

OLLAMA_ENABLED = os.getenv("OLLAMA_ENABLED", "False").lower() == "true"
OLLAMA_URL = os.getenv("OLLAMA_URL", "http://localhost:11434/api/generate")
OLLAMA_MODEL = os.getenv("OLLAMA_MODEL", "llama3.2:3b")
OLLAMA_TIMEOUT_SECONDS = int(os.getenv("OLLAMA_TIMEOUT_SECONDS", "12"))

Family-Centric Experience

WellNest is designed around the idea that wellness is not an individual responsibility, but a shared family experience. Instead of isolating users, the platform brings family members into the loop in a meaningful and supportive way.

Family members can view summarized wellness insights such as mood patterns and risk signals, allowing them to stay aware without invading privacy. Based on these insights, they can send care_pulses, messages, or encouragement to support the user.

These interactions are captured within the system, creating a continuous feedback loop of care and engagement. Rather than reacting after problems arise, WellNest enables early support and preventive intervention.

This approach transforms wellness tracking into a collaborative and emotionally supportive experience, making it easier for users to stay consistent and feel supported in their daily routines.

Challenges We Faced and How We Solved Them

Building WellNest involved challenges in handling real-world user data and maintaining performance. Noisy and inconsistent inputs were addressed using validation rules and lightweight NLP techniques. To support near real-time processing, we optimized queries with indexing and efficient filtering. Evolving data structures were managed using MongoDB’s flexible document model. We also organized the backend using modular Django apps to ensure scalability and clarity. Overall, these solutions helped create a robust, adaptable, and smooth user experience.

Meet Nesti: Your Wellness Companion

Nesti is WellNest’s built-in chatbot that provides real-time, emotionally supportive guidance for users. Designed as a gentle companion, it helps users reflect on stress, mood, sleep, and routines through simple conversations.
It uses a layered AI approach—LLM responses when available, fallback text-insight logic, and safe deterministic replies—to ensure reliability.
Integrated with check-ins, AI insights, and family support features, Nesti delivers context-aware interactions. Overall, it turns WellNest into an interactive, preventive wellness platform.

What’s Next for WellNest

WellNest can be further enhanced by improving its AI capabilities and expanding its real-time interaction features. Future updates may include more advanced models for deeper emotional insights, along with improved personalization of recommendations.

We also plan to introduce notification and intervention workflows, allowing timely alerts when risk levels increase. Enhancing explainability in AI insights and tracking long-term wellness trends will provide users with better clarity and guidance.

Additionally, moving towards a mobile-first experience and integrating with wearable devices can make wellness tracking more seamless and continuous.

Wrapping Up

WellNest is more than a wellness tracking app—it is a family-centric, AI-assisted platform focused on preventive care and emotional well-being. By combining daily check-ins with intelligent insights and family support, it helps detect early stress signals and enables timely intervention.

With Django as the orchestration layer and MongoDB as the flexible data backbone, the system efficiently handles dynamic wellness data. Lightweight AI and ML components provide meaningful, context-aware insights without adding complexity.

Overall, WellNest promotes a collaborative and proactive approach to wellness, where care is shared and sustained through technology.

GitHub Repository: https://github.com/srikiran1905s/pfsd_MAIN.git
Demo video:

DEV Community: Pippalla Sri Kiran