Architecture Overview

What is Aito?

Aito is a predictive database — a new category of database that combines traditional data storage with built-in machine learning. Unlike bolting ML models onto existing databases or building custom ML pipelines, Aito provides predictions as a native database operation.

Core Components

Storage Layer

Your data is stored in tables, similar to a relational database. You define a schema and upload data via the API or console.

Indexing Layer

When you upload data, Aito automatically builds probabilistic models that capture relationships between fields. No manual feature engineering or model training required.

Query Engine

The query engine handles prediction requests, similarity searches, and data retrieval. Each prediction includes:

• The predicted value
• A confidence score (0-1)
• Explanation of which factors influenced the prediction

Deployment Options

Sandbox

Ideal for experimentation and development. Shared infrastructure with logical data separation. Limited to 5,000 API calls/month.

Dev & Production

Dedicated compute resources in AWS EU (Ireland). Full tenant isolation. Scales based on your data volume and query patterns.

Enterprise On-Premise

For organizations requiring data to stay within their own infrastructure. Aito can be deployed on your AWS account or private cloud. Contact us for details.

How Predictions Work

Unlike traditional ML where you:

1. Extract features
2. Train a model
3. Deploy the model
4. Update when data changes

With Aito, you:

1. Upload data
2. Query for predictions

The predictive index updates automatically as you add or modify data. No retraining, no deployment pipelines, no ML expertise required.

Scaling Considerations

• Data volume: Aito handles millions of rows efficiently
• Query latency: Typical prediction queries return in 10-100ms
• Concurrent requests: Rate limits apply per plan (see Reliability)
• Index updates: Near real-time as data is added

Integration Patterns

Aito fits into your existing architecture via REST API:

Common patterns:

• Sync on change: Push data to Aito when your database updates
• Batch sync: Periodic bulk updates from your data warehouse
• Direct integration: Query Aito in real-time for user-facing predictions

Security Architecture

See our Security & Compliance page for details on:

• Tenant isolation
• Encryption
• Authentication
• Data residency

Questions?

For architecture discussions or custom deployment requirements, please contact us.