SATURDAY, MAY 2, 2026FR

SIMON LECHEVALIER

Software Development, Full-Stack & AI-Augmented Engineering for production-oriented web, mobile, automation, and AI-enabled products.

I design cross-platform products with a pragmatic engineering stack: Next.js, React Native, FastAPI, Python, Docker, cloud-ready APIs, and AI-assisted development workflows. My strength is turning LLM tooling into concrete delivery leverage: faster architecture exploration, safer refactors, stronger documentation, automated checks, and higher-quality implementation loops.

I am looking for engineering opportunities where I can combine full-stack delivery, AI tooling mastery, automation, and product sense to ship reliable software faster without lowering code quality.

The Developer

CONNECT

AI Development Workflow

METHOD

Context Engineering

Frames tasks with architecture notes, constraints, acceptance criteria, repo context, and verification targets before asking AI tools to implement.

Agentic Coding

Uses Claude Code, Codex, MCP servers, plugins, and specialized prompts to explore codebases, generate plans, implement changes, and review diffs.

Automation Loops

Builds shell and Python scripts to repeat setup, debugging, validation, file operations, environment checks, and development workflows across machines.

Quality Control

Treats LLM output as a draft: runs lint/build/test checks, inspects generated code, rewrites unclear logic, and keeps changes reviewable.

Featured Works

VIEW ALL PROJECTS

Jarvis Voice Assistant

ProblemBuilding a mobile AI assistant requires low-friction voice capture, provider-safe backend access, real-time response flow, persistent memory, and a UX that degrades cleanly when voice services fail.
SystemExpo React Native client, FastAPI backend, REST/WebSocket voice endpoints, Deepgram STT/TTS, Groq LLM generation, Kokoro ONNX fallback, PostgreSQL knowledge base, Redis working memory, Docker Compose infrastructure, and latency-conscious provider orchestration.
OutcomeDelivered a complete low-latency voice pipeline that can answer a spoken question in about 2 seconds, proving the architecture was designed around fast capture, transcription, reasoning, synthesis, playback, and graceful text fallback.
#ReactNative#Expo#FastAPI#Python#Deepgram#Groq#PostgreSQL#Redis#Docker

ConvAI Speech Analysis Agent

ProblemExtracting useful insight from speech needs more than transcription: audio ingestion, analysis orchestration, structured outputs, and reusable agent workflows must work together reliably.
SystemPython-first AI agent pipeline for speech analysis, LLM-assisted reasoning, structured architecture notes, scriptable automation, and API-oriented integration points for future product surfaces.
OutcomeBuilt a foundation for conversational/speech intelligence workflows that demonstrates AI orchestration, automation discipline, and the ability to move from raw voice input toward actionable analysis.
#Python#LLM#SpeechAI#Agents#Automation#APIs#Architecture

StreamArena

ProblemWeb3 prediction products often lose mainstream users through fragmented onboarding, wallet complexity, and unclear real-time feedback.
SystemPrediction workflow, smart-contract interaction layer, unified web interface, TypeScript front end, PostgreSQL-backed data model, and latency-conscious API design.
OutcomeDesigned a Web2-friendly path into Web3 prediction mechanics while keeping the product architecture maintainable and ready for real-time interactions.
#Next.js#React#TypeScript#PostgreSQL#Web3#Solidity#APIs

AI Engineering Workflow

ProblemLLM tools can create speed but also noise when prompts, context, verification, and architecture decisions are not controlled.
SystemClaude Code, Codex, MCP servers, plugin workflows, architecture markdown files, shell automation, test/build loops, Dockerized environments, and cross-OS development on macOS, Linux, WSL, and Windows.
OutcomeUses AI as an engineering multiplier: faster code exploration, sharper implementation plans, repeatable automation, stronger documentation, and reviewable delivery instead of unstructured code generation.
#ClaudeCode#Codex#MCP#Plugins#PromptEngineering#Shell#Docker#CrossPlatform

Professional Experience

AI-Augmented Full-Stack Developer

Aug 2022 - Current
Freelance

Designs and ships web, mobile, and automation systems with AI-assisted engineering workflows. Uses Claude Code, Codex, MCP tooling, architecture notes, shell scripts, Docker, and structured verification loops to accelerate delivery while keeping code reviewable, maintainable, and aligned with product goals.

Full-Stack Developer

Jun 2025 - Nov 2025
Teclis Scientific

Revamped a legacy Wix system into a modern Next.js platform over a six-month engagement. Introduced CMS-driven maintenance, expanded multilingual support from 3 to 10 languages, improved SEO architecture, structured content workflows, and helped turn a difficult-to-maintain marketing site into a scalable international platform.

Web3 Front-End & Integration Developer

Jan 2022 - Dec 2025
Independent Web3 Projects

Built and explored decentralized product architectures across wallet flows, smart-contract interaction patterns, token-gated experiences, prediction mechanics, and Web3 libraries. Focused on reducing onboarding friction for Web2 users, connecting blockchain state to usable interfaces, and designing systems where security, transaction clarity, and UX feedback remain understandable.

NOW BUILDING

ACTIVE
Jarvis

Voice-first AI assistant with an end-to-end STT/LLM/TTS pipeline optimized for low latency, typically answering in about 2 seconds from spoken question to response.

Progress75%

Tools & Tech

FOR HIRE, SKILLED IN:
  • AI Coding Systems95%
  • Claude Code / Codex95%
  • MCP & Plugins90%
  • Prompt Engineering92%
  • TypeScript88%
  • HTML / CSS / JavaScript92%
  • Web3 Libraries62%
  • Python / FastAPI75%
  • React / Next.js90%
  • React Native / Expo82%
  • Docker / Compose70%
  • Shell / Bash / PowerShell88%
  • Cloud APIs82%

Engineering Strengths

  • Architecture-first implementation
  • AI-assisted refactoring
  • Cross-platform debugging
  • API integration and provider boundaries
  • Developer automation
  • Technical documentation
  • Product-minded UX decisions

Systems & Tooling

  • macOS, Linux, Windows, WSL
  • Bash, Zsh, PowerShell, Python scripts
  • Docker, Docker Compose, local infra
  • Xcode, iOS Simulator, Android Emulator
  • Git, GitHub, CI/CD awareness
  • PostgreSQL, Redis, cloud API providers

Additional Notable Experience

Shopify & E-Commerce

Practical knowledge of Shopify storefront constraints, conversion-oriented pages, product presentation, checkout friction, SEO basics, and merchant-focused content workflows.

Crypto Ecosystem

Strong understanding of wallets, tokens, DeFi culture, on-chain UX, Web3 communities, market narratives, smart-contract interaction patterns, and the gap between crypto-native and mainstream users.

Product Design

Comfortable shaping product flows, information hierarchy, interface clarity, onboarding, and feature scope so technical systems remain understandable to real users.

3D & Blender

Basic Blender and 3D workflow knowledge useful for asset preparation, visual prototyping, spatial thinking, and collaboration around WebGL or immersive interface ideas.

Languages

  • French[NATIVE]
  • English[BUSINESS LEVEL]
  • Japanese[DAILY CONVERSATION]

Hobbies

LanguagesCosmologyGeopoliticsFinanceTravelJapanese CultureProgrammingAIVideo GamesMusic