Aito Grocery Store Demo

🚀 Try the Live Demo → demo.aito.ai

Experience AI-powered grocery shopping with personalized recommendations, smart search, and predictive cart filling!

Aito.ai is a predictive database, that provides real time predictions, recommendations and statistics via instant SQL-like queries without separate train step. The goal of the system is to drastically reduce the effort needed to create basic machine learning applications.

The Aito.ai demo highlights 13 production-ready ML features that can be build extremely quickly using the predictive database capabilities. Each live feature comes with screenshots, code samples and tutorials.

Try It Now

# Test the API instantly (no signup required)
curl -X POST https://shared.aito.ai/db/aito-demo/api/v1/_predict \
  -H "X-API-Key: bvss2i2dIkaWUfBCdzEO89LpPNhqjD" \
  -H "Content-Type: application/json" \
  -d '{
    "from": "products",
    "where": {"name": {"$match": "milk"}},
    "predict": "tags"
  }'

What I Built

1. 🎯 Dynamic Recommendations

{
  "from": "impressions",
  "where": {
    "context.user": "larry",
    "product.id": { "$and": [{ "$not": "10" }, { "$not": "15" }] }
  },
  "recommend": "product",
  "goal": { "purchase": true },
  "select": ["name", "id", "tags", "price"],
  "limit": 5
}

→ Implementation | Use case guide | 🚀 Live Demo

2. 💡 Intelligent Autocomplete

{
  "from": "contexts",
  "where": { 
    "queryPhrase": { "$startsWith": "mil" },
    "user": "larry" 
  },
  "get": "queryPhrase",
  "orderBy": "$p",
  "select": ["$p", "$value"]
}

→ Implementation | Use case guide | 🚀 Live Demo

3. 🔍 Smart Search with Personalization

{
  "from": "impressions",
  "where": {
    "product": {
      "$or": [
        { "tags": { "$match": "milk" } },
        { "name": { "$match": "milk" } }
      ]
    },
    "context.user": "larry"
  },
  "get": "product",
  "orderBy": {
    "$multiply": [
      "$similarity",
      { "$p": { "$context": { "purchase": true } } }
    ]
  },
  "select": ["name", "id", "tags", "price", "$matches"],
  "limit": 5
}

→ Implementation | Use case guide | 🚀 Live Demo

4. 🏷️ Automated Tag and Category Predictions

// Predict tags
{
  "from": "products",
  "where": { "name": "Rye bread" },
  "predict": "tags",
  "exclusiveness": false,
  "limit": 10
}
// Returns: ['gluten', 'bread'] (filtered by $p > 0.5)

// Predict category
{
  "from": "products",
  "where": { "name": "Rye bread" },
  "predict": "category",
  "limit": 1
}
// Returns: { feature: "101", $p: 0.83 }

AI-powered product catalog management predicts tags, category, and price for new products → Implementation & src/13-product-predictions.js | Use case guide | 🚀 Live Demo

5. 📝 Smart Cart Autofill

{
  "from": "visits",
  "where": { "user": "larry" },
  "predict": "purchases",
  "exclusiveness": false,
  "select": ["$p", "$value"]
}

// Filter results: hit.$p >= 0.4 (40%+ confidence) → Implementation | Use case guide | 🚀 Live Demo

6. 🗣️ NLP Text Classification

{
  "from": "prompts",
  "where": { "prompt": "Which payment methods do you provide?" },
  "predict": "answer"
}

// Returns: sentiment: 'negative', category: 'user_experience' → Implementation | Use case guide | 🚀 Live Demo

7. 📊 Statistical Relationship Discovery

{
  "from": "visits",
  "where": { "user.tags": "club-member" },
  "relate": "purchases"
}

// Returns: lift scores showing what club members buy more → Implementation | Use case guide | 🚀 Live Demo

8. 📄 Automated Invoice Processing

{
  "from": "invoices",
  "where": { "Description": "AWS Cloud" },
  "predict": "Processor",
  "select": ["$p", "Name", "Role", { "$why": { "highlight": true } }]
}

→ Implementation | Use case guide | 🚀 Live Demo

9. 📈 Product Analytics Dashboard

[
  {
    "from": "impressions",
    "where": { "purchase": true },
    "relate": { "product": "42" },
    "select": ["lift", "related"]
  },
  {
    "from": "visits",
    "where": { "purchases": { "$has": "42" } },
    "relate": "user.tags",
    "select": ["lift", "related"]
  }
]

→ Implementation | Use case guide | 🚀 Live Demo

10. 🤖 Customer Assistant

{
  "from": "impressions",
  "where": {
    "product.tags": { "$match": "gluten-free bread" },
    "product.price": { "$lte": 5 }
  },
  "orderBy": { "$p": { "$context": { "purchase": true } } }
}

// "Find gluten-free bread under $5" → Implementation | Use case guide | 🚀 Live Demo

11. 🔧 Employee Assistant

{
  "from": "impressions",
  "where": { "purchase": true, "context.week": "2024-03" },
  "get": { "$group": "product.name", "$stats": { "sales": { "$count": true } } },
  "orderBy": { "sales": -1 }
}

// "What are our top selling products this week?" → Implementation | Use case guide | 🚀 Live Demo

12. 📊 Model Quality

{
  "test": { "$index": { "$gt": 90 } },
  "evaluate": {
    "from": "impressions",
    "where": { "context.user": { "$exists": true } },
    "predict": "purchase"
  },
  "select": ["accuracy", "meanRank", "meanMs", "cases"]
}

// Returns: accuracy: 0.87, meanRank: 2.3, meanMs: 45 → Implementation | Use case guide | 🚀 Live Demo

13. 💰 Price-Demand Analytics

{
  "from": "price_history",
  "where": {
    "product_id": "milk_organic",
    "day_of_week": "Friday",
    "is_weekend": false
  },
  "estimate": "sale_price",
  "model": "regression",
  "select": ["estimate", "why"]
}

// Returns: optimal price with demand forecast and profit maximization → Implementation | Use case guide | 🚀 Live Demo

🚀 Quick Start

Click to expand installation instructions

# Clone and run locally
git clone https://github.com/AitoDotAI/aito-demo.git
cd aito-demo
npm install
cp .env.example .env  # Includes working demo credentials
npm start

The app opens at http://localhost:3000. No API key setup required - uses public demo instance.

For your own Aito instance:

# Edit .env with your credentials
REACT_APP_AITO_URL=https://your-instance.aito.app
REACT_APP_AITO_API_KEY=your-api-key

🎯 Technical Highlights

Performance: 90K impressions dataset, <100ms query latency, no cold starts Schema: Proper Aito format with linked tables (users → visits → contexts → impressions → products) Real Data: 134 users, 42 products, 90,087 interaction records Architecture: React frontend + Aito.ai backend, fully responsive

📖 Deep Dive

• Blog Post: Complete technical walkthrough
• Use Case Guides: Detailed implementation guides for each feature
• Assistant Integration: How to build AI assistants with Aito
• Data Model: Schema design and relationships

🤝 Why This Matters

Replace months of ML pipeline development with SQL-like queries. No feature engineering, no model training, no deployment complexity. Just queries that return predictions.

Traditional ML: Feature pipelines → Model training → Serving infrastructure → Maintenance Aito.ai: Query → Prediction ✅

---

Built in a weekend to solve a real client problem. The entire codebase is open source - MIT licensed.