Agent Research Company is the practice of Alexander Izak Levesque, an independent advisor on agentic trading systems based in Quebec and serving clients globally. The work covers the deployment, strategic analysis, and systemic-risk implications of multi-agent systems in financial markets — from prop desks and family offices standing up agent-mediated execution, through banks and asset managers reading the competitive map, to the central banks, securities regulators, and exchanges now writing the supervisory frame around all of it.
The fifteen-year arc began in cryptocurrency. Active in Bitcoin from 2010, I co-organized the first sustained Bitcoin meetup at the Bitcoin Embassy in Montreal — the first dedicated Bitcoin physical hub on Earth — and worked across the institutional crypto cycle that followed: market structure, custody, on-chain forensics, and the incident-response practice that grew out of repeated wallet-compromise and exchange-failure work. I have submitted written opinions to the Ontario Securities Commission on cryptocurrency policy and have advised counsel and counterparties through enforcement and post-incident remediation.
The technical work has spanned market-data systems and market-data spend analysis for hedge funds and trading firms; custodial technologies and hardware hardening; and CTO-level leadership at VR Generation in Dubai — a behavioural-data and virtual-reality research practice using advanced simulated environments to study human decision-making — and at Key Advisors, a law firm where I led the technology function. In Paris I contributed to the APAD standard, working with museums, ministries, artists, and collectors on the intellectual-property infrastructure for digital art. The throughline is the same problem in different surfaces: how institutions absorb new technical primitives faster than their governance can describe them.
The present concentration is agentic trading: the architecture, the regulation, and the systemic implications. The April 2026 Bank of England Financial Policy Committee record names the risk plainly — that as multiple firms deploy agentic systems built on common models and harnesses, those systems may synchronise and amplify market moves, with the potential for a coordinated, machine-driven flash crash. At the same time, hedge funds are now standing up entire analytical functions — sentiment, fundamentals, quantitative analysis, risk management — as fleets of cooperating agents; one Sydney-based fund operates with no human analysts at all. The gap between what the buy side is shipping and what the supervisory frame is reading is the gap this practice exists to close.
A solo practitioner is the right shape for that gap. Independence is structural; the work is written; one engagement at a time.
