1.5 billion smartphones

can now run generative AI on-device. The customer relationship is moving with them.

[methodology]

44×

uplift in activation effectiveness when personalisation runs on the customer’s device.

[methodology]

The Platform

DNA3 
One orchestration layer for Device Native: Attributes, AI and Agents.

Running where your customer is. Designed by you.

DataSapien DNA³ is the context layer your customer platforms have been missing. It collects, structures and acts on first-party signal where it is freshest, on your customer’s own device, and makes that intelligence composable with your existing CDP, campaign, loyalty and analytics tools. Attributes, AI models and agents in one orchestration layer. Your team composes. The platform runs.

A1

Attributes

One live profile per device, no new silo.

Behavioural, transactional, declared and inferred signals merge into a single live customer profile on the device. Every tool downstream reads from the same source. Your CDP and warehouse stay where they are.

A2

AI

Models that run where the data lives.

Personalisation models, scoring, classification and ranking execute natively on the device. No round-trip to the cloud. No latency tax. No data lake to defend.

A3

AGENTS

Orchestrate action across your tools, on your terms.

Trigger workflows, compose interactions, and route actions across your existing systems through a no-code interface. You define what acts, where, and under what conditions.

Developers

A native SDK for real-time, on-device customer context.

Drop DataSapien into your iOS, Android, React Native or Flutter app. A small language model, context store and agent runtime, all on-device, with a typed API and a sync layer you control.

// 1 · Install via Swift Package Manager
//   https://github.com/datasapien/ios-sdk

import DataSapienSDK

// 2 · Configure once, in your App entry point
DataSapien.configure(apiKey: "YOUR_API_KEY")

// 3 · Query lived context, on-device
let ctx  = await ds.context(.session)
let reco = await ds.recommend(in: ctx)
Speed
Context query latency, median, on-device
< 50 ms
Median latency for on-device context queries, measured on iOS and Android.
Footprint
SDK footprint, includes context store and model environment
~20 MB
Real-world footprint at current build. Includes the on-device runtime, context store, and bundled small language model.
Availability
Uptime, deployed devices, trailing 90 days
99.97%
Uptime across deployed devices, measured via on-device observability.

Measured on SDK v1.6, production builds, trailing 90 days.