Skip to content
YK Développement

Work

Complete products, from concept to production.

Five full-stack projects spanning web, mobile, desktop and R&D, designed, architected and built end to end. Three are shown in detail, two remain private.

Scroll to explore

Mequi

In production

Business SaaS connecting riders and equestrian coaches.

8 features · 43 technologies

  • Web SaaS
  • Marketplace
  • Payments

Key features

  • Shared coach/rider scheduling with booking, availability and careful time-zone handling.
  • Geolocated coach search on an interactive map, with PostGIS clustering.
  • Progress tracking: goals, work plans, session reports and a training journal.
  • Horse management: profiles, photo galleries, shared access and a dedicated dashboard.
  • Subscriptions and payments via Stripe Connect, with billing and generated PDF documents.

Results

  • A complete SaaS in production, from first contact with a coach through to payment.
  • End-to-end business coverage: over 35 backend modules running in production.
  • Turborepo
  • npm workspaces
  • Zod partagé
  • React 18
  • TypeScript
  • Vite
  • Tailwind
  • shadcn/ui
View the case study

Ceryo

TestFlight

All-in-one mobile super-app, smooth and offline-first.

8 features · 32 technologies

  • Mobile
  • iOS
  • Android

Key features

  • Offline-first Batch Snapshot architecture: one screen, one endpoint, instant perceived load.
  • Multi-channel real-time via Action Cable: chat, presence, daily plan and learning.
  • AI assistant and semantic search powered by pgvector embeddings.
  • Unified native health integrations: HealthKit on iOS and Health Connect on Android.
  • Rich modules: planning, learning, budget, media, atlas/travel and a TipTap editor.

Results

  • Apple-grade mobile experience: 60 fps animations, haptics and a canonical art direction.
  • Full-stack mastery end to end: React Native, real-time Rails and AI.
  • React Native 0.76
  • Expo 52
  • TypeScript
  • React Navigation
  • Reanimated 3
  • Gesture Handler
  • FlashList
  • TanStack Query
View the case study

GodMode

Active

Native desktop cockpit for local projects and AI models.

5 features · 12 technologies

  • Desktop
  • Rust
  • Tauri

Key features

  • Project dashboard: per-service start/stop/restart with a green/amber/red health dot.
  • Automatic detection of services by parsing manifests (package.json, Cargo.toml and others).
  • Live panel: a view of listening ports and running processes, even outside GodMode.
  • MLX control: kill AI models individually or globally from the menu bar.
  • Live log tailing, health checks, global shortcuts and macOS notifications.

Results

  • A tool in daily use: start, stop and supervise every local project in one gesture.
  • Modular Rust architecture (registry, process manager, scanners) with no per-project hardcoding.
  • Rust
  • Tauri 2
  • tokio
  • serde
  • rusqlite
  • regex
  • tracing
  • IPC natif
View the case study

Alpha

Private

Real-time autonomous system — SSOT, runtime invariants, extreme robustness.

5 challenges · 26 technologies

  • Backend
  • Real-time
  • R&D

Private project — I share the stack and the engineering challenges, never the business logic.

Engineering challenges

  • Unified SSOT architecture: a single source of truth driving identical code across modes.
  • Runtime invariants checked continuously, able to trigger an emergency stop on detected drift.
  • Dual-pool data architecture (control plane / session plane) with strict routing.
  • Extreme robustness: idempotence, reconciliation, drift detection and safe cold-boot.
  • Testing discipline: AST regression scans and property-based tests to guarantee invariants.

What I can share

  • Alpha is a real-time autonomous system that I rebuilt entirely on a clean foundation. The project is private: I don't detail its functional logic, but it embodies the level of engineering rigor I carry into all my work, systematic root-cause fixing, zero cosmetic patches and verifiable safety nets.
  • Python 3.12
  • FastAPI
  • asyncpg
  • PostgreSQL 17
  • Alembic
  • Pydantic 2
  • pydantic-settings
  • structlog
View the project

Morpheus

Private

Multi-agent AI orchestration engine, provider-agnostic.

5 challenges · 5 technologies

  • AI
  • Multi-agent
  • R&D

Private project — I share the stack and the engineering challenges, never the business logic.

Engineering challenges

  • Multi-agent orchestration: specialized roles, self-improvement loops and per-agent memory.
  • Provider-agnostic LLM integration: seamless switching between local MLX and cloud models.
  • Long-running autonomy: extended execution, incident recovery and robustness.
  • Strict environment isolation: dedicated port, home directory and env prefixes to coexist.
  • Specialized pipeline: analysis, decision, audit and meta-improvement stages chained together.

What I can share

  • Morpheus is an autonomous multi-agent engine, the R&D predecessor from which I distilled the architectural patterns later reused in my real-time projects. Private as well: I share its approach and engineering challenges, not its functional details.
  • Python
  • orchestration multi-agents
  • LLM provider-agnostic
  • MLX local + cloud
  • isolation stricte
View the project

The next product might be yours.

Describe what you need and get an estimate in minutes.

Get my quote
Work — YK Développement