HI, I’M ODILON
Odilon Buisson, embedded systems and AI automation engineer
Ideas. Systems. Connections.
An ESME Sudria embedded systems engineer, I design AI agents, products and automations that actually get things done.
AI-generated portrait
01 / The project network
6 STATIONS · 6 LINES↳ Each station leads to the project's case study.A map of connections, not a timeline.
L1VoiceTheCallAgentAI agents that answer on behalf of a small business: on the phone, on WhatsApp and on its website.Read the case study 
L2AppsWeChat Mini ProgramsEnd-to-end design and development of WeChat Mini Programs for the Chinese market.Read the case study 
L3AI toolingPersonal AI agents, deployed and managedSetting up an always-on, secure and controllable personal AI agent for an individual or a small team, demonstrated on my own agent in production.Read the case study 
L4EnterpriseAI-assisted recruitingClient engagement: a high-volume hiring pipeline supported by AI, where every decision is still made by a recruiter.Read the case study 
L5CreationCode-driven video production workshopA production pipeline for vertical video: AI-generated shots, voice-over, animated captions and code-based rendering.Read the case study 
L6ResearchScientific research: radar, MATLAB and AIThesis on radar-based drone detection: signal processing in MATLAB, machine learning in Python, and AI agents that keep the pipeline reliable.Read the case study
Conceptual illustrations · AI-generated
03 / BEHIND THE NETWORK
From embedded systems
to AI agents.
I’m Odilon Buisson, an engineer who graduated from ESME Sudria in 2026, majoring in embedded systems, and an entrepreneur. The same year I am completing a master’s degree at the Beijing Institute of Technology, where I am finalising my thesis on radar-based drone detection.
I taught myself AI alongside my coursework: documentation, personal projects, then clients. I specialised in agents for businesses — the ones that act rather than chat: answering the phone, finding prospects, following up, updating a CRM with no retyping. Engineering school taught me to make a system hold up in the real world; AI gives it autonomy, as long as the rules stay in the code and a human stays in the loop.
Education
- ESME SudriaEngineering degree (master’s level) · major in embedded systemsGraduated in 2026
- Beijing Institute of TechnologyDual-degree master’s · thesis on radar-based drone detection2026, thesis being finalised
- University of Wollongong, AustraliaExchange semester: automation, robotics, Java programming, managementSix months
- Self-taughtApplied AI: agents for businesses, automation, agent-assisted developmentAlongside my studies, since 2025
Expertise
Embedded systems
- Engineering training at ESME Sudria, major in embedded systems; automation and robotics at the University of Wollongong
- Radar and signal processing: W-band (77 GHz) FMCW chain, micro-Doppler signatures, MATLAB
- Drones: professional remote pilot certified by the French civil aviation authority (DGAC) since 2022, swarm flight for the inspection of port equipment
- From physical measurement to decision: features, machine learning, statistical validation
AI agents for businesses
- Voice agents: bookings and appointments taken over the phone (availability, booking, cancellation, confirmation by e-mail and SMS), outbound calls from a list, a report of every call in the CRM
- Messaging and web: WhatsApp agents and website agents, assistants that answer from the company’s own documents
- Prospecting: finding companies, enriching and qualifying contacts, verifying addresses, e-mail sequences, scheduled competitor monitoring
- LinkedIn and content: replies to messages drafted by AI and approved in one click in Slack, LinkedIn posts and blog articles published on a schedule
- Documents: quotes, invoices and contracts generated as PDFs from a simple request in Telegram or Slack, replies to e-mails that come with a PDF, OCR, customer follow-ups by e-mail and SMS
- Management assistants: a main agent, driven from Telegram or Slack, delegates to sub-agents (calendar, e-mail, CRM, quotes); adapted for restaurants, real estate and joinery
- More than 80 n8n workflows built, around fifty of them migrated to a self-hosted instance
- Business rules in the code and the database, never in the prompt alone; human approval for sensitive actions
Right now
- Thesis: radar-based drone detection ↗Retraining the models on corrected data splits, then adapting them to real W-band data.
- Dronaparc × CLIMA drone swarm for inspecting port structures: one drone per task, centimetre-level positioning, AI detection of cracks and corrosion, a summary by an expert engineer. The aim: fewer aerial platforms and rope teams, and shorter equipment downtime.
- TheCallAgent ↗AI agents for small businesses, on the phone, on WhatsApp and on the web. Read the case study
- WeChat Mini Programs ↗A bilingual prototype tested automatically in the official simulator, being prepared for the Chinese market.
- Personal AI agents ↗An always-on agent deployed in the cloud with its own control console, ready to be replicated.