Cyril Moron

Technical Lead for the core internal framework serving as the software foundation for Combat Management Systems (CMS). My role focuses on ensuring the robustness of legacy systems while preparing architectural evolutions.

• Framework & Middleware Architecture: Design and maintenance of the distributed CMS core foundation based on 3 critical pillars:
  • • Data Distribution: Real-time data exchange (DDS/DLRL) and distributed configuration (JGroups).
  • • Resilience & Availability: Implementation of fault tolerance mechanisms at both machine and process levels.
  • • Inter-system Communication: Remote method invocation (Control/Command) via CORBA.
• Expert Support: Acting as the final line of defense for high-complexity technical issues (distributed system crashes, memory leaks, network latency).
• DevOps & Release Modernization: Leading the software factory transformation: migrating legacy Jenkins setups to an industrial GitLab CI / Nexus stack. Ownership of release cycles and versioning.
• Python Technical Authority: Defining the Group’s normative technical reference for Python (coding standards, tooling, industrialization) within a Defense environment.

Software integration and system configuration for Submarine Combat Management Systems, focusing primarily on the Scorpène-class (Brazilian PROSUB program).

• Software Integration: Full-stack assembly of the CMS, handling OS configuration, network layers, and software startup sequences.
• Domain Expertise: Integration and validation of critical subsystems including Tactical Situation and Weapons.
• Performance & Stability: Analysis and optimization of system constraints (scalability, boot time, operational readiness).
• International Tech Transfer: Conducted technical training and support for the Brazilian Navy (Scorpène program), ensuring their autonomy on software assembly.

Development and integration for Submarine Combat Systems (Barracuda and Scorpène classes), with a strong focus on critical weapon systems and hardware interfacing.

• Weapon Launch Sequences: Specialized in designing and implementing robust software state machines for critical weapon launch sequences, including SM39 (Exocet), F21 (Torpedo), and MdCN (Naval Cruise Missile).
• Equipment Interfacing: Deep analysis of manufacturer Interface Control Documents (ICDs) to develop the software modules responsible for parsing and managing complex network exchange formats with the MdCN.
• Subsystems Engineering: Core backend development for Tactical Situation and External Sensors subsystems.

Consumer web platform combining a configurable pace calculator and a search engine over the French Athletics Federation database (940,000 athletes, official records, weekly scraper). Paired with MyPacer Club, a B2B module for athletics club managers and volunteers that turns the collection of weekend race results into automated communication.

• Full-stack architecture: Python API (FastAPI, uv), SvelteKit web, FFA scraper, MyPacer Club dashboard. Independent sub-packages, containerized, deployed to staging and production via Docker Compose and GitHub Actions.
• Agent-native pivot: the DecaSaaS commercial pipeline (FFA club prospecting, follow-ups, season renewals) is operated by Hermes agent (Nous Research) through a dedicated CLI. Python/Typer designed as an execution surface for an LLM: structured JSON output, remote/local modes, scoped tokens, lazily-loaded Markdown operator skills.
• Stack: Python (FastAPI, uv), SvelteKit, PostgreSQL, Docker, GitHub Actions, Hermes agent.

Instances of personal AI agents powered by OpenClaw, deployed and secured on an isolated virtualized Linux infrastructure.

• Custom domain skills: technology watch, application development, sport & health monitoring, multi-agent team for brainstorming technical topics.
• Multi-model layering: frontier model as the main orchestrator for its strong agentic capabilities and autonomy, delegating specialized tasks to lighter LLMs (typically DeepSeek). Capability / latency / cost arbitration.
• Security & industrialization: isolation in a dedicated KVM VM, secrets managed outside the workspace, Telegram access restricted to paired contacts, idempotent Ansible deployment, systemd user services.

Coherent family of Python CLIs over daily self-tracking tools: official APIs (Strava, Withings) or reverse-engineered private mobile APIs (Coros, Yazio). Dual output: rich tables for human use, structured JSON for AI agent consumption.

• strava-cli: running activities, segments and statistics via the official Strava API.
• withings-cli: body composition (scale) and ECG recordings (BPM Core) via the official Withings API.
• coros-cli: sleep metrics extracted from the Coros app.
• yazio-cli: nutrition tracking over the private Yazio API.