TinyML & Edge AI: On-device inference, model quantization, embedded ML, ultra-low-power AI for microcontrollers and IoT devices.

iOS + Android app that runs local LLMs on-device + routstr cloud LLMs for anonymous inference

Flutter starter example app to get started with NobodyWho, a library designed to run LLMs locally and efficiently on any device.

Mobile AI: iOS CoreML, Android TFLite, on-device inference, ONNX, TensorRT, and ML deployment for smartphones.

Production Android AI with ExecuTorch 1.0 - Deploy PyTorch models to mobile with NPU acceleration and 50KB footprint

Custom llama.cpp fork with character intelligence engine: control vectors, attention bias, head rescaling, attention temperature, fast weight memory

High-performance Android SDK for on-device LLM inference (GGUF). Privacy-focused, offline-first, and powered by llama.cpp with a clean Kotlin Coroutines API.

Neural acoustic echo cancellation for Apple platforms using CoreML — Swift package with 128/256/512-unit DTLN-aec models

Real-time SAM2 segmentation on edge devices - 40x faster C++ inference with ONNX Runtime for iOS/Android deployment

Ad generation via offline LLMs with on-device inference, optionally managed by a self-hosted CMS.

The Private Agent OS — search files, run AI agents, connect to 10,000+ tools via the complete protocol stack (MCP, AG-UI, A2UI, A2A). Zero cloud. Zero telemetry. On-device inference.

Swift wrapper for Apple's BNNS graph API — run compiled CoreML models (.mlmodelc) on CPU with zero-copy buffer management

Run small LLMs directly on your device, no cloud computing needed.

React Native SDK for local LLM inference and on-device AI on iOS and Android.

Open source Node.js runtime for local LLM inference, on-device AI, and private model execution.

Web JavaScript SDK for local LLM inference with WebGPU and on-device AI.

WebGPU runtime core for local LLM inference, on-device AI, and client-side model execution.

On-device inference engine for Apple silicon

Offline plant disease diagnosis system powered by MobileNetV3-Large and TensorFlow Lite — 38 disease classes, 14 crop species, ~5.58ms inference on-device. Built with Flutter & Python.

Deep technical writing on edge AI, on-device inference, llama.cpp, GGML, and mobile AI engineering