Direct to the Point
Herman Cheung spent six and a half years at O’Melveny & Myers before joining Covenant full time as its direct investments attorney. His background in corporate finance and venture investments fills a gap Covenant is actively building toward.
A Different Kind of Practice
When Herman joined Covenant, he brought something the team didn't yet have: deep experience in corporate finance, spanning secured and unsecured credit facilities, convertible debt and venture financings from pre-seed to Series E and beyond. That experience shapes how he thinks about what Covenant is building.
"A tool is as good as it's designed for," he says. "If it's purposefully made for a specific use case, designed by someone who has actually done that type of work, it's going to be far more useful."
Covenant has expanded beyond its initial focus on fund document review into direct investments and credit, which is Herman's home territory. The ability to bring the same AI-powered approach to that category of work, informed by someone who has lived inside it, is what his role is built around.
Where AI Changes the Work
Venture financing is a particularly good use case for AI-powered legal review, and the reason comes down to standardization. Early-stage financings are typically done on SAFEs, Simple Agreements for Future Equity, a form developed by Y Combinator that has become widely adopted across the market. When a company is ready for a priced round, the National Venture Capital Association releases model documents that serve as the market standard. Because these forms are so established, the negotiating points are largely consistent from deal to deal: protective provisions, pre-emptive rights, anti-dilution protections, etc.
That consistency is exactly what makes AI effective here. Covenant has built a standard playbook that reviews venture financing documents against market standard documents, as well as customized client playbooks for investors who have specific points they ask for in every deal. The platform reviews the full document set and surfaces what matters without a lawyer having to start from scratch each time. The turnaround is two business days, meaningfully faster than the timeline most investors are used to.
“AI is eliminating a lot of the repetitive elements of the drafting and negotiation process by reviewing against largely standardized forms and flagging when standard protections are missing,” Herman says. “This frees lawyers up to focus on provisions that are bespoke to the deal.”
Speed as a Structural Advantage
The point Herman returns to most isn’t efficiency but rather adaptability. AI is moving fast enough that the ability to adopt new tools quickly has become a meaningful competitive differentiator. Large institutions face structural headwinds here: dozens of practice groups, different workflows, different client bases, and every new tool becomes a negotiation. By the time consensus is reached, the technology has often moved on.
“They want to bring in AI, but the rollout is slow at a large established institution,” Herman says. “AI moves so fast that being slow is a real disadvantage.”
At Covenant, a new model or capability can be evaluated and integrated immediately. “We’re very purposefully focused in our products,” he says. “That tool can be applied much more nimbly. And in this environment, where AI is advancing so fast that every week brings new capabilities, that speed is necessary.”
Herman’s bet in joining Covenant full time is that being small and purpose-built is the better position to be in when the pace of change is this fast. “We cannot ignore this very powerful tool,” he says. “I really wanted to be involved early, working at that intersection between law and AI.”
