Cognitive Industries is a small independent lab: human engineering, amplified by a fleet of very opinionated AI agents. The surface area is unreasonable on purpose. A from-scratch operating system in Rust, heterogeneous compute that targets GPUs and quantum backends, self-improving reasoning engines, real-time vision, and the products you've already scrolled past. We ship research-grade systems and production apps with the same standard: do the hard thing properly, or don't do it.
Systems & Low-Level
A from-scratch Rust operating system with capability security and deterministic replay, and an intent runtime whose formal proofs run in microseconds.
→ DominionOS · WeftScript
AI & Machine Learning
Models trained from scratch, self-improving reasoning engines, evolutionary architecture search, plus the shipped products that lean on them.
→ ChazzAI · TTHUNK · EvoArch
Computer Vision
Real-time GPU pipelines: licence plates through extreme degradation, FaceID with persistent tracking, fire and smoke detection.
→ CogVis · FireSmokeDetector
Security & Cryptography
Non-LLM vulnerability scanning, a self-playing cyber range, encrypted-at-rest stores, and licensing backed by proof-of-work.
→ Unslop · MXCyberSLaM · HumanIO
Web Platforms & SaaS
Multi-tenant products with payments, admin consoles, and mobile builds from one codebase. Several of them are live right now.
→ ChazzAI · MatchPoints · WebSwarm
Cloud & Infrastructure
A Cloudflare-native platform (Workers, D1, queues, SSO, a metered AI gateway) runs every app on this page from one account system.
→ HiveMind platform · 13 production Worker deploys
Quantum & Heterogeneous Compute
A provider-agnostic IR that plans each piece of work onto CPU, GPU, or quantum hardware with cost estimation across real QPU providers.
→ FluxionOS · 204 tests, 91% coverage
Media & Streaming
Multi-tenant live streaming with automated capture, encoding, and scoreboard overlays; in-browser file tooling where nothing is uploaded.
→ Boomerang · WebTools Ultima
A sample of the research and systems work behind the scenes. Not all of it ships as a product; some of it just had to exist.
HETEROGENEOUS COMPUTE
FluxionOS
"Write once, run on any accelerator, including QPUs." A provider-agnostic IR that plans whether each piece runs on CPU, GPU, or quantum hardware, with cost estimation across IBM Quantum, AWS Braket and D-Wave. 204 tests, 91% coverage.
Python · Qiskit · Braket · Ocean
AUTONOMOUS REASONING
NARS Cognitive Engine
A self-improving engine where a symbolic NARS core reasons autonomously, escalates to LLMs only when stuck, then compiles those rescues into permanent knowledge. Proved 6/7 theorems and rediscovered 3 physics laws, 50% LLM-free.
Python · PyNARS · SymPy · PySR
CONFIDENTIAL COMPUTE
VideoGen
Text-to-video on an 8GB GPU where frames never touch disk as plaintext: a cascade-encrypted, deniable in-memory vault with anti-capture viewing. It can optionally burst to confidential H200 compute.
Python · Diffusers · XChaCha20
TEST-TIME COMPUTE
TTHUNK
A 9.6M-parameter model trained from scratch that beats far larger models once paired with test-time search and verification, a 7× lift at scale. The negative results are documented as carefully as the wins.
Python · PyTorch
COGNITIVE ARCHITECTURE
ManyBrains
Tests whether intelligence emerges from many small specialised models behind a learned router: Mixture-of-Experts meets Global Workspace Theory. Routers learn domain from loss alone, and reasoning paths vary in length.
Python · PyTorch
EVOLUTIONARY ML
EvoArch
Evolves all seven components of an LLM stack at once (preprocessing, tokeniser, architecture, training, inference, post-processing, representation), with a UCB1 bandit choosing what to improve so gains compound on a shared genome.
Python · OpenEvolve
