Artificial intelligence is only useful if you know how to ask the right question. “I think your prompt has to be something very well thought of and drafted,” says Lara Merjane, corporate counsel at JBS USA. “One of the main instructions I gave is to not make anything up.” Merjane has built Microsoft Copilot agents embedded in her company’s contract playbooks – a practice that traces directly to her years at The Ross Firm Professional Corporation, a boutique spread across six small Ontario towns where the firm used AI to redline contracts well before most practices had considered it.
Licensed in Brazil, Ontario, California, and Colorado, Merjane arrived at JBS USA in December 2025 with a foundation built across civil and common law jurisdictions, a background in agricultural and corporate-commercial work, and hands-on experience with contract review automation that the role would demand from day one.
From rural Ontario to global procurement
At a smaller practice, lawyers “deal [with] the complex parts of the matters earlier on” – something she found harder to access during a later stint at a large firm. After completing an LLM at Osgoode Hall Law School and qualifying in Ontario, Merjane joined The Ross Firm Professional Corporation, where she worked, with that large-firm stint in between, for approximately five years. The firm’s six small-town Ontario offices meant she was handling corporate-commercial matters alongside agricultural work: acquisitions for chicken companies, farm agreements, expropriation of farm land, and deals for a client base that didn’t look like any big-firm roster.
The firm’s technology culture was as formative as its client mix. Under the guidance of Quinn Ross, the firm adopted AI for contract review earlier than most practices, equipping lawyers with a tool that could generate redlines and attach suggested revision comments at the click of a button. The connection to her current approach is direct: her heavy early use of the technology is what drove her to start building playbooks and Copilot agents. The move to JBS USA came through a referral from an HR contact – Merjane wasn’t actively looking – and the fit was immediate. The company’s Brazilian origins, strong Brazilian leadership presence, and agricultural focus aligned with her background, and her long-term plan to relocate somewhere warmer accelerated faster than expected.
Inside JBS USA: scope and mandate
JBS USA is part of a group of more than 300 companies and is part of JBS N.V., the world’s largest protein producer, with beef, pork, poultry, and leather operations spanning North America and beyond. Merjane’s mandate covers all global procurement contracts across the group’s entities, international logistics – air freight, ocean shipping, trucking, transloading, dry brokerage – IT agreements, some M&A, and responsibility for JBS Canada commercial matters. She also handles legal matters in which Brazil and the US overlap, or in which JBS USA is a party to a global arrangement. The legal team in Greeley, Colorado, where JBS USA is headquartered, numbers approximately 20 lawyers working in person, and she reports to the general counsel.
The food industry’s regulatory density is built into every contract. Standard agreements include extensive representations, warranties, and protections, with a dedicated in-house regulatory department available to address specialized questions. When volume spikes and multiple departments simultaneously hit legal with urgent requests, external resources absorb the overflow. “That juggling of priorities, of urgencies is… the main time management concern for any in-house lawyer,” Merjane says, “because we never know when two or three departments are going to have an urgency at the same time.”
AI in practice: agents, playbooks, and guardrails
The Copilot agents Merjane has built at JBS USA are trained on the company’s own playbooks, not general knowledge. “I built Copilot agents that have the contract review functionality,” she says. “So I put our playbooks into an agent.” The system flags deviations from the company’s standard positions; anything outside those positions gets verified against her knowledge of what is acceptable. For actual legal research, she uses LexisNexis or Westlaw – not Copilot. The distinction matters: AI processes and applies known positions; established legal research platforms generate independent legal analysis.
Two practical rules anchor her approach to AI governance. First, the prompt must be precise, and the tool must be told explicitly not to fabricate. Second, configure it to ask for clarification rather than fill gaps with invented answers. “And also I ask that it asks me questions back if it’s not sure of something,” she says.
Cross-border practice and legal culture
Merjane finds that risk allocation and commercial contract principles translate fairly well across jurisdictions. The more meaningful differences are cultural. “I think some cultures are much more used to relying on a statute,” she says – South American counterparties trained in civil law systems often struggle with the judicial discretion and precedent-based reasoning that characterizes common law contracting in North America. “They’re more risk-averse, and they really want something in writing,” she adds.
On the Canada-US divide, her read is direct: “Americans are a lot more direct than Canadians. But other than that, I find the two systems very similar.” The sharper adjustment, in her experience, sits between Latin America and North America – not between the two North American systems.
Understanding the business as the core mandate
Her advice for lawyers considering an in-house move comes back to one foundational requirement. “The main homework that we have to do is to understand the business, understand all the business units, understand what you’re selling, what you’re purchasing and what are the goals of the leadership so you can adapt your advice to achieve… those goals,” Merjane says. The practical outcome of that understanding is advice framed in terms of probability and mitigation – presenting risks, gauging their likelihood, and giving leadership what they need to make informed decisions. That business judgment, she says, is built through experience and learning from senior colleagues, and it transfers across countries, jurisdictions, and practice settings.