Applied AI for Scientific Computing & Embodied Intelligent Digital Twins
Feaxr LLC bridges artificial intelligence with scientific computing to understand, simulate, and optimize complex physical systems. Rather than building general-purpose models, we focus on high-impact domains including energy science, additive manufacturing, and real-time physical diagnostic tools.
Transforms mobile devices into diagnostic labs. Features a patch-based MLP-Mixer with MoE and linear gated attention for instant battery health, thermal performance, and aging forecasting.
Voice-activated copilot and assistive navigation tool. Built on a 100% private, on-device architecture for multi-step task execution, contextual memory, and screen reading.
Minimalist AI skills launcher designed to streamline prompt engineering, store reusable workflows, and execute context-rich prompts across ChatGPT, Claude, Gemini, or directly inside the app.
A unified entrance interface for AI interaction. Acts as the core platform foundation for Feaxr's future scientific and engineering software suites.
On-device optical inspection app designed to detect hidden camera lenses and infrared reflections using real-time Computer Vision.
Visual assistant for real-time optical character recognition, multi-language translation, object scanning, and secure barcode parsing.
Web-based Lithium-ion simulation platform paired with an embedded LLM assistant. Designed for researchers to simulate cell behaviors, configure models, and interpret physical parameters naturally.
Research workflow converting smartphone photographs into simulation-ready 3D meshes. Combines AI 3D reconstruction with parallel finite element analysis to compute residual stress and additive manufacturing deformations.
Every line of code is engineered to solve validated physical and engineering challenges.
Rooted in rigorous physical principles, numerical accuracy, and validated models.
AI designed to elevate and collaborate with human domain expertise, not replace it.
Privacy-first architectures engineered for local computation and low latency.