Causal Foundry SDK - Cross-Platform AI/ML Data & Intervention Engine

[SDK]

Tech Stack

Kotlin, React Native, React, Javascript, Typescript, Java, iOS, Swift

Organization(s)

Causal Foundry

Resources

SDK Docs

Causal Foundry SDK - Cross-Platform AI/ML Data & Intervention Engine

Project Details

The Challenge

Causal Foundry's AI/ML platform needed a reliable, lightweight bridge between partner applications and its reinforcement learning engine, capable of capturing behavioral data, delivering adaptive interventions, and working seamlessly across Android, iOS, React Native, and Web, without becoming a burden on partner apps or their users' devices.

The Solution

As the lead engineer, I designed and built the Causal Foundry SDK from the ground up across all four platforms, ensuring a modular, high-performance data ingestion and intervention engine that serves as the core gateway between partner applications and the CF AI/ML platform. The SDK captures behavioral signals, delivers personalized nudges, and auto-logs user responses, powering the reinforcement learning feedback loop that drives adaptive, personalized experiences across partner domains.

What I Built

Behavioral Data Ingestion: Captures predefined and custom events including app lifecycle, screen-level usage durations, and domain-specific user interactions across e-commerce, e-learning, healthcare, and loyalty programs

Intervention Engine: Delivers timely, AI-driven nudges in multiple forms: push notifications, in-app messages, and content-based/UI nudges; with automatic response logging (engaged, dismissed, or blocked) feeding directly back into the ML platform

Android SDK: Built using Kotlin, leveraging WorkManager for efficient background task scheduling and event ingestion, optimized for performance and battery efficiency. Published on Maven Central

iOS SDK: Built from scratch in Swift, using BGTaskScheduler for native, resource-aware background data logging. Published on CocoaPods and Swift Package Manager

React Native SDK: Bridging layer with a TypeScript wrapper over the native Android and iOS SDKs, centralizing network and data management while accelerating release cycles. Published on NPM

JavaScript Web SDK: Extended platform reach to partner websites, supporting event ingestion and intervention nudge delivery for web-based applications

Modular Architecture: Segmented into domain-specific modules (Core, E-Commerce, E-Learning, Loyalty, Payments, Community Health) to minimize package size and keep integrations lean and purposeful

Impact

Deployed across global partner platforms spanning e-commerce, healthcare, and loyalty programs enabling Causal Foundry's AI/ML engine to deliver personalized, adaptive interventions at scale in some of the world's most underserved markets.