Skip to main content
Patent BP-084 · Mobile AI SDK

Enterprise AI
on Every Device

Auto-generated typed mobile SDKs for React Native, Flutter, Swift, and Kotlin. Offline WASM inference. SSE streaming. SHAP explainability. All 19 AI platforms. One SDK call to rule them all.

4
Mobile platforms
React Native · Flutter · Swift · Kotlin
0ms
Offline inference latency
WASM cached on device — no server call
19
AI namespaces
All BrainPredict platforms in one SDK
BP-084
Patent pending
Paris Convention — Apr 16 2026

How Mobile AI Works

From API call to App Store in 4 steps

01

Auto-Generated SDK

Call GET /api/v1/sdk/mobile/{platform} and receive a fully typed, production-ready SDK. React Native TypeScript, Flutter Dart, Swift, or Kotlin — each with all 19 BrainPredict AI platform classes pre-wired.

02

WASM Offline Inference

The first time the app launches, it downloads a compact WASM module containing the tenant's GBM tree structure. Subsequent predictions execute entirely on-device — zero network round-trip.

03

SSE + SHAP Streaming

When online, predictions stream in real-time via Server-Sent Events. Each frame includes the prediction, SHAP top-3 factors, Merkle hash, and confidence interval — automatically displayed as explainable AI cards.

04

Resilient Retry + Auth

Built-in 3-attempt exponential backoff, OAuth2 Bearer token management, and platform-native secure storage (MMKV / Hive / Keychain / EncryptedSharedPreferences). Enterprise-grade out of the box.

SDK Features

Offline-First Architecture

WASM inference runs without internet. Predictions cached in platform-native storage. Syncs when back online.

Typed Platform Namespaces

FinanceAI, SupplyChainAI, HRAI, RiskAI... all 19 platforms as typed classes with IntelliSense/autocomplete.

Real-Time SSE Streaming

IB-triggered Server-Sent Events stream predictions as they happen. Zero polling required. Battery-efficient.

SHAP on Mobile

Explainability data included in every response. Display "Why this prediction?" cards compliant with EU AI Act Art.13.

NLQ Voice Interface

Built-in OTTO multilingual NLQ integration. Business users ask questions in natural language — 80+ languages.

Post-Quantum Ready

Kyber-768 key exchange and Dilithium-3 request signing built into all SDK network calls. Quantum-safe from day 1.

Ready to ship AI to mobile?

One API call generates your complete typed mobile SDK — production-ready in minutes.

Request SDK Access