Coder, Maker, Entrepreneur, Epicurean Taoist, Quantified & Sovereign ❤️
📍 EU-Africa-Gulf-India Corridor or Remote | 💼 Available for Strategic Advisory, Fractional CTO & Frontier Architecture

Founder of @bradtech (DeepTech Agronomy & Industrial IoT) and @Quatrain (Enterprise Master Data Management & Finite State Machines). Creator of the Tycho open-source autonomous infrastructure ecosystem.

I believe the future of artificial intelligence must not be centralized into monolithic cloud silos. Real intelligence is sovereign, frugal, and edge-native.

flowchart LR
    subgraph Edge ["The Sovereign Edge (0ms Local-First)"]
        User["Human Mind & Bio-Data"] <--> Storage["OKF Markdown & Local DB (IndexedDB / SQLite)"]
        Storage <--> EdgeAI["Edge AI & Embedded SLMs (On-Device Intelligence)"]
    end

    subgraph Cloud ["Opportunistic Cloud (Zero Lock-In)"]
        EdgeAI <--> Sync["Background Multi-Cloud Sync (PostgreSQL / K8s / S3)"]
        Sync <--> HeavyAI["Asynchronous High-Order LLMs"]
    end

Data must belong irrevocably to the user or tenant. In Modaka, applications perform 0ms local reads and writes via client-side databases (IndexedDB, SQLite, local disk) and synchronize opportunistically in the background. If the network drops, your second brain and tools work without interruption.

No proprietary database lock-in. OKF stores knowledge as flat, human-readable Markdown bounded by structured YAML headers. It provides deterministic semantic structures that both human beings and autonomous AI agents can query through progressive disclosure.

Pushing inference directly to user devices, robotics controllers, and IoT gateways using Small Language Models (SLMs) and on-device neural embeddings. Heavy cloud models are reserved strictly for asynchronous reasoning, guaranteeing data privacy, minimal energy footprint, and instantaneous responsiveness.

LLMs generate possibilities, but physical reality demands determinism. By coupling AI agents with strict Finite State Machines (FSM) and Domain-Driven Design (DDD), we build mission-critical systems that are intelligent yet provably reliable for industrial IoT, robotics, and medical telemetry.

Passionate advocate for personal health metrics sovereignty, biological telemetry, and longevity science. Building local-first, privacy-preserving tools to model, track, and optimize physiological well-being without vendor lock-in.

Exploring the new frontier: mission-critical edge computing, real-time robotics firmware (C++ / PlatformIO / RTOS), satellite and terrestrial LoRaWAN telemetry, and fail-safe state machines designed for extreme, high-reliability environments.

Designing zero-friction, sovereign cloud systems: Kubernetes ARM64/AMD64 clusters, declarative GitOps with ArgoCD, automated ACME/Let's Encrypt TLS orchestration, and rootless Podman container environments (Tycho).

Languages       : TypeScript (Strict/No Any), Modern C++ (Embedded/Firmware/Robotics), Python, Bash/POSIX
Hardware & IoT  : ESP32 / ARM Microcontrollers, PlatformIO, LoRaWAN (ChirpStack/Basics Station), Sensor Networks
Edge & AI       : Local-First (IndexedDB/SQLite), Small Language Models (SLMs), On-Device Embeddings, OKF v0.1
Cloud & Infra   : Kubernetes (ARM64 & AMD64), Terraform, Traefik v3, Scaleway, AWS, Podman Rootless
Architecture    : Domain-Driven Design (DDD), Finite State Machines (FSM), 15-Year Horizon Maintainability

All public and enterprise architectures in my portfolio adhere to strict, long-term engineering standards:

• 🛡️ LLM Agent Development Principles & GitFlow Protocol (AGENTS.md)
• 🔒 Zero Hardcoded Secrets & Fail-Fast Infrastructure-as-Code
• 🧪 Zero Untested Code & Test-Driven Development (TDD)
• 🌍 15-Year Horizon & Clean Modular Architecture

• 💼 LinkedIn: linkedin.com/in/crapougnax
• 🐦 Twitter / X: @crapougnax
• 🏢 Organizations: @tycho-ops • @bradtech • @Quatrain