# Olivier Couthaud, CV Builds AI agent systems and says what was and was not tested. Starts agentic QA at Third Axis in November 2026. Aix-en-Provence, France · US citizen · oliviercouthaud.com · github.com/forkei · linkedin.com/in/olivier-couthaud ## Experience ### AI platform engineer, agentic QA, Third Axis AI (From Nov 2026) Part-time contract. Using an AI-employee product the way a business owner would, and building the harness of automated agents to test it at volume and find what breaks. ### Contract developer, Adapter Group (May – Sep 2026) Built agent prototypes on a knowledge-graph platform's API for a client. Details confidential. ### Contract developer, Metropolis World (Apr – May 2026) Built the first phase of an AI-character platform in a four-week, fixed-scope contract: a chess-playing character that talks, analyses the game and is meant to feel like a person you are playing against. Gave it memories (generated from a backstory) and emotions, building on earlier work on agent memory and emotion; also the character data model and a web interface. ### President, RIT AI Club (Sep 2025 – May 2026) Ran the club for the 2025–26 year: hundreds of members on the roll, 10 to 50 attending each session in person. Guest speakers, events and workshops. Set up a platform for students to learn AI at their own pace, aimed at certifications, as a base for later officers to build on. ### Software tutor, Wyzant (Jun 2025 – Mar 2026) Built at least six client projects while teaching each client how and why they worked, including a retrieval-based assistant for an enterprise client and a finance research agent. ## Projects ### WorldGraph, investigation agent (2026) Stack: TypeScript, PostgreSQL. Status: early, code not public. More: https://oliviercouthaud.com/worldgraph Agent that searches the web, reads pages and queries public records, then builds an evidence graph where each claim is stored with the quote and page it came from. Same-name entities stay separate until evidence links them; every merge is logged and reversible. Quotes are not checked against the page, so some are paraphrases. Single user; the code is not public. ### Balatro, Rust copy of the game and an agent (2026) Stack: Rust, PyTorch. Status: in progress, code not public. More: https://oliviercouthaud.com/balatro Rust copy of the game replayed against the real game in two passes: 727 actions of simple play, then 1,165 across packs, skips and all 28 bosses. No real mismatches (two timing differences judged harmless). Agent trained from scratch on about a billion moves (about 33 hours, one rented GPU): wins about 2 in 3 White-stake games, Red deck. The main agent wins none on Gold, the hardest stake; a separate Gold-trained version wins 2 to 5 in 100. ### VOS, multi-agent assistant (Sep 2025 – Feb 2026) Stack: FastAPI, RabbitMQ, PostgreSQL, Weaviate, Gemini, Flutter. Status: prototype. More: https://github.com/forkei/VOS-public Built the SDK every agent runs on, the API gateway, the primary agent, most of the specialist agents and the phone-call system (Twilio gateway). The voice gateway is shared work; the app was designed by Roman Slack. A primary agent hands work to specialist agents through message queues; text chat and phone calls. A prototype with no automated tests. ### Liaison, messaging between AI agents (2026) Stack: CRDT, signed certificates. Status: prototype, no revocation yet. More: https://github.com/forkei/liaison Peer-to-peer messaging between AI agents with no central server: a signed message log (CRDT), certificates tying each agent to its owner, and encrypted private messages. Group posts are signed, not encrypted. Tried between a Windows laptop and a Raspberry Pi over the public internet. Prototype: keys stored unprotected on disk, no certificate revocation. ### HealControl, IBM Dev Day hackathon (team of four) (Jan 2026) Stack: Python, MCP (FastMCP). Status: demo, one toy program. More: https://github.com/hl9082/healcontrol Wrote the MCP server (12 actions) that lets an AI assistant run tests, patch a bug, check the fix and open a pull request. The demo runs end to end on one toy program, the only thing it has been run on. ### Fault-tolerant quadrotor control (2026) Stack: JAX, MuJoCo MJX, PPO. Status: simulation only. More: https://github.com/forkei/fault-tolerant-quadrotor Reimplemented the published SimpleFlight tracking recipe (PPO) on MuJoCo MJX and JAX: 2,048 parallel environments, about 70,000 steps per second on one RTX 4070. First policy, figure-eight at normal speed, 3 seeds: 0.040 m mean error against 0.028 m in the paper; no crash in 21 benchmark rollouts (7 trajectories, 3 seeds). Found that the paper's evaluation code starts the figure-eight at T/4, not t = 0; a fault-trained policy looked 2.5 times worse than the first one until both used the same start. Added RMA-style adaptation to one rotor at 70% thrust (privileged teacher, history encoder): 0.057 m healthy (0.040 m before fault training) and 0.081 m with the fault at normal speed; fast runs with the fault crashed on 2 of 3 seeds. Separate depth-camera obstacle policy: 92% of fair scenes collision-free, 52% with the fault, and its tracking missed the 0.10 m target (0.17 m). The repository has the first policy and the fault-training setup; the encoder and obstacle policy are not on GitHub yet. ### Gaussian splats and House23D (2026) Stack: COLMAP, hloc, nerfstudio, Blender. Status: one room checked. More: https://oliviercouthaud.com/splats Built the glue from phone video to Gaussian splat around COLMAP, hloc and nerfstudio: frame selection, camera poses, training on a rented RTX 4090, about $0.40 per training run. Switching pose method took a night scene from about 10% of photos placed to 98%. House23D: rebuilt a bedroom as a scaled Blender model from phone video and photos, checked by overlaying renders on the photos. The errors are not measured systematically. ### HeartbeatID, ECG identity network (2026) Stack: PyTorch, PhysioNet ECG-ID. Status: public data, one seed. More: https://oliviercouthaud.com/heartbeatid 0.1M-parameter network on PhysioNet ECG-ID: 92.4% top-1 on 10 held-out people, tested on a different day. One seed and a small test set, on public recordings; a Galaxy Watch app is built and was run once, with no results from the watch yet. ### PicoTalon, 3D-printed RC plane (2026) Stack: 3D printing. Status: assembled, not flown. More: https://oliviercouthaud.com/picotalon 905 mm plane printed in pieces and assembled, 400 g with battery; waiting on its electronics, it has not flown. ### Engine sound simulator (2026) Stack: Rust. Status: desktop app live, browser version in progress. More: https://oliviercouthaud.com/enginesim Rust simulation of gas flow through an engine's intake and exhaust, turned into sound live instead of played back from recordings. ## Education Rochester Institute of Technology, New York, Sep 2024 – Jan 2026. Computer science, accelerated BS/MS. Not completed: left in January 2026 to build full time. ## Skills - Languages: Python, TypeScript, Rust, Dart (Flutter) - Systems and tools: FastAPI, PostgreSQL, Weaviate, RabbitMQ, Docker, Linux, GCP, MCP, Gemini and Anthropic APIs - ML: PyTorch, JAX, PPO - Human languages: English and French, both native