I design secure, deterministic, and business-aware AI systems for real-world workflows. My background in institutional asset management brings a strong discipline around control, validation, governance, and risk — which I now apply to multi-agent systems, RAG pipelines, MCP-based integrations, and Web3 / Fintech products.
I build AI systems that do more than generate answers: they verify information, call tools safely, validate transactions, and operate with traceability in mind.
- Specialization: Multi-Agent Systems, RAG, MCP, Secure AI Workflows, Web3 & Fintech
- Current focus: AI architecture, trustworthy automation, agentic systems in production
- Open to: 24-month apprenticeship / junior AI engineer roles / solutions architecture opportunities
Python · TypeScript · React · AWS · Bedrock AgentCore · MCP · RAG · Agentic Workflows · FastAPI · Docker · Solana · GitHub Actions · SQL
I work at the intersection of AI engineering and operational reliability, with a strong interest in systems where data integrity, transaction validation, and human trust matter.
| Project | What it does | Why it matters |
|---|---|---|
| novamart-multi-agent-support | A multi-agent support system for an e-commerce workflow | Shows orchestration, role separation, and specialist agent patterns |
| grounded-research-agent | A grounded RAG agent that answers only from trusted sources | Demonstrates source-grounded reasoning and refusal behavior |
| my-gecko-buyer | A buyer agent that validates transactions before signing | Focuses on trust, verification, and secure execution |
| support-chatbot-bedrock-agentcore | AI support chatbot architecture with evaluation | Brings product thinking and system-level AI design |
| agentcore-customer-support-agent | End-to-end agent with tools, knowledge retrieval, and orchestration | Illustrates practical integration patterns |
| one-step | A local AI productivity assistant running in-browser | Explores privacy-first, lightweight AI experiences |
After 10 years in institutional asset management, I moved into AI engineering because I wanted to build systems that combine operational rigor with intelligent automation.
That experience shaped my way of thinking:
- validation over speed
- traceability over opacity
- governance over improvisation
- trusted execution over hype
I bring that same financial discipline to AI architecture, especially in domains where correctness and accountability are critical.
- AWS AI & ML Scholars
- Future AWS Agent Engineer Nanodegree
- Strong interest in production-grade AI systems, multi-agent collaboration, and secure automation
I started coding to build a simple website for my own business. That quickly turned into a deeper interest in problem-solving, systems thinking, and software architecture.
Two things became clear:
- I enjoy being stuck on a technical problem until it becomes clear.
- I like building systems that do real work, not just demonstrations.
That was the turning point.
- 📚 My Building Library — my technical notes, writing, and learning journey in English and French
- ✍️ dev.to/mialy333 — articles and technical reflections
- 🔭 Currently exploring: full-stack fundamentals, Docker/Kubernetes, and a multi-agent project built from first commit to production-ready architecture
I am seeking a 24-month apprenticeship / junior AI engineer position where I can contribute to the deployment of trusted agent-based systems in production.
I’m especially interested in:
- Multi-agent systems
- RAG + knowledge-grounded AI
- MCP and tool integration
- AI for finance, operations, and digital services
- Web3 / Fintech applications with strong validation and security requirements
Open to feedback, collaboration, and conversations around AI architecture, agent systems, and trustworthy automation.
If one of these projects helped you, a ⭐ would mean a lot.