Interdisciplinary Creative Technologist, Artist and Developer
Vincent Charlebois
Software developer, artist, and agentic interaction designer
Website · GitHub · LinkedIn · Email · Art
I build deployable AI systems, distributed infrastructure, product-focused digital tools, and browser-native artworks. My current focus is designing and developing interactions with autonomous systems: making agent actions observable, controllable, and useful in real-world workflows. I work across Go and Rust backends, terminal interfaces, decentralized protocols, and high-fidelity web applications.
Professional Experience
Software Developer — Fly.io
| Remote | July 2026–present |
- Developing data persistence and core storage systems in Go and Rust for Fly.io’s Computers for Agents infrastructure.
- Working on durable state for hardware-isolated Linux environments designed for long-running AI agents and arbitrary code execution.
Member-worker — Hypha Worker Co-operative
Montréal, QC | 2022–July 2026
- Lead developer of RooLLM, a tool-using AI system deployed in the co-op’s organizational workflows.
- Designed modular agent architecture integrating local LLM inference, structured tool execution, and API orchestration.
- Built backend services (Node.js and Rust integrations) supporting decentralized storage and distributed workflows.
- Implemented citation-grounded retrieval systems and human-in-the-loop approval flows.
- Deployed and maintained self-hosted GPU inference infrastructure.
- Integrated AI systems into production dashboards with TypeScript and Svelte.
Rust Backend Contributor (Contract) — OpenArchive
| Remote | 2023–2024 |
- Contributed to save-dweb-backend, a peer-to-peer archival backend for the Save mobile app.
- Developed Rust storage and networking components within an architecture using Veilid for peer discovery and routing and Iroh for blob storage and integrity verification.
- Focused on resilient citizen-media preservation without a central point of control.
Technical Stack
- Languages: Go, Rust, TypeScript/JavaScript, Python, Solidity
- AI systems: Tool-using agents, RAG, local inference, structured outputs, evaluation
- Agent infrastructure: Trace-based observability, durable memory, safety controls, sandboxing, CLI/TUI design
- Backend & protocols: Node.js, REST APIs, distributed systems, AT Protocol, Veilid, Iroh
- Infrastructure: Docker, Linux, CI/CD, self-hosted inference, decentralized storage
- Frontend: Svelte, interaction design, performance-focused UI systems
Selected Technical Projects
v100
Go runtime for building, running, and studying autonomous coding agents.
Implements interchangeable reasoning strategies, schema-bound tools, execution budgets, durable memory and traces, replay, evaluation, and operator safety controls.
Doxa & Nego
Browser-native artworks investigating how people understand and participate in autonomous systems.
Make agent memory, state, errors, and multi-agent negotiation legible as interaction spaces.
Anaphor
Rust terminal tool that selects and gathers relevant context from the web, URLs, local files, or standard input.
Produces source-grounded answers from natural-language questions without requiring users to choose tools manually.
spores.garden
Decentralized publishing client built on the AT Protocol.
Frontend architecture for lexicon-extensible content systems.
Previous Roles
- Web Specialist & Multimedia Coordinator — McGill University Library (2021–2022)
- Production & Operations Manager — Fablab inc (3D printing startup) (2019–2020)
- Project Manager — Reforestation operations across Canada (2012–2018)
Education
Master in Advanced Ecological Buildings & Biocities
IAAC / UPC — 2019
BFA, Intermedia & Cyberarts
Concordia University — 2012