Shailaja-poojari/ai-last-mile-delivery-optimizer

AI-Powered Last-Mile Delivery Optimizer

This project is now officially open for open-source contributions

✨Walmart Sparkathon Focused

✅ Problem Statement

• Retailers and e-commerce giants like Walmart face significant challenges in last-mile delivery, especially during high-demand events (festivals, emergencies, traffic congestion).
• Delays, inefficient routing, high carbon emissions, and poor real-time coordination with delivery agents are major bottlenecks.

---

✅ Solution
Built an intelligent, modular, and scalable dashboard that:

• Optimizes delivery routes in real-time
• Supports Emergency Mode, 🎉 Festival Mode, and future AI toggles
• Works even in offline mode with local caching
• Tracks agent GPS live
• Calculates fuel and CO₂ savings
• Includes a smart AI chatbot assistant for delivery managers

---

This is not a customer-facing app.

✨ It’s a plug-and-play AI-powered dashboard for Walmart’s internal last-mile delivery operations, designed to help managers reduce delays, costs, and carbon emissions while improving visibility.

⭐ What Sets Us Apart ⭐

• A truly plug-and-play system — easy to integrate without overhauling Walmart’s existing tools
• Works offline, supports live GPS, and is extremely modular
• Not just analytics — it acts, by intelligently optimizing delivery plans and helping reduce operational chaos
• Supports real-time AI-based delay predictions, risk scoring, and sustainability tracking
• Built with scalability in mind — works across zones, cities, and third-party partners
• Even saving ₹10–15 fuel cost per delivery, when scaled to 1 lakh+ daily deliveries, can save Walmart over ₹1 crore every month — purely from smarter routing, consolidation, and delay mitigation .
• Our solution makes that possible — in a plug-and-play, AI-powered, and GPS-aware dashboard.

---

Key Features

• AI-Powered Optimization
  Predicts delivery delays using zone, distance, and seasonal factors
  Computes risk scores for each order
• Smart Consolidation
  Suggests grouped deliveries for similar localities and short distances
• Displays consolidation summary + AI-generated suggestions
• Emergency / Festival Mode
  On one click, apply optimization to minimize high-risk zones (e.g., Koramangala)
• Easy to extend with new modes (Rain Mode, Traffic Mode, Partner Priority Mode)
• Real-Time GPS Tracking
  Uses live agent GPS to improve visibility
• ETA Countdown & Risk Monitoring
  Each order has live ETA countdown, delay severity badge, and zone-based risk flagging
• Sustainability Metrics
  Calculates fuel and CO₂ savings from optimized routes
• 🧑‍💻 Smart Chatbot Assistant
  Provides help and suggestions for dashboard usage
  Answers common delivery/logistics questions (React-component based)
• Offline Mode
  Works seamlessly when internet is unavailable using localStorage caching
• Delivery Agent App (Companion)
  Mobile-friendly interface for agents to view assigned orders, optimized routes, and update delivery status
  Helps reduce confusion, delays, and syncs live with the main dashboard

---

Why This Solution is Plug-and-Play
⚡Companies can adopt it without rewriting existing tools
No hard dependency on internal APIs — can connect via Supabase, REST, or partner APIs
Uses simple JSON/mockOrder format that real systems can map to easily
⚡Extremely Modular Code
Each function and feature is decoupled
AI modes are plug functions like applyEmergencyMode() or applyRainyMode()
⚡No Vendor Lock-in
Backend is Supabase (PostgreSQL) → Can switch to any backend
Frontend is React + Tailwind → Works with any modern stack

---

Scalability & Upgradeability
⚡ Can Handle 1000s of Deliveries per Hour
Efficient rendering (React + Tailwind UI)
Grouping and filtering logic already optimized
Uses minimal external dependencies (low overhead)
⚡Add New Delivery Partners Easily
Integrate via API or CSV → maps to existing format
UI adjusts dynamically based on zone, partner, distance
⚡Add New Cities / Zones
Our system uses zone fields — adding a new city is plug-and-play
No core logic changes required
⚡Add More AI Modes Anytime
Already supports toggles for Emergency / Festival
Can easily add: Rain Mode, Peak Hour Mode, Partner Priority Mode etc.
✅ Just plug new logic file → toggle UI → done.

---

✨ We’re also looking forward to making this solution more scalable for real-world operations
by integrating live delivery systems, expanding to more cities, and adapting to dynamic business needs at scale.

---

🛠 Tech Stack
Frontend: React.js + Tailwind CSS
Backend: Supabase (PostgreSQL + API)
Hosting: Vercel (or Netlify/AWS/GCP)
Map: Static coordinates for demo, GPS via browser
AI: Rule-based predictors + delay scorers (can upgrade to ML models)

---

Final Pitch Statement
“Our solution is AI-powered, offline-resilient, GPS-aware, and scalable across cities, partners, and use cases — giving Walmart a plug-and-play delivery optimizer ready for real-world operations.”

---

Built For:
E-Commerce
Hyperlocal Delivery
Logistics Tech
Walmart Sparkathon

---

👩‍💻 Author
Shailaja Poojary
Creator & Maintainer of the AI Last-Mile Delivery Optimizer

---

HOW TO RUN
'''bash
Clone the repo
npm install
npm run dev

---

Ensure you have .env set up for Supabase keys (or use mock mode).