How Goozo brought its contract standard into one working system

The challenge

Goozo ran a CRM before Lexnus. It was bought to keep contracts in one place and it did that job, sort of. NDAs and MSAs went in, versions accumulated, and the documents stayed retrievable but spread over multiple places (documents, agreements, e-signatures) with a high manual process to find and retrieve them. The limits sat one layer above the documents, in the policy those documents were meant to express. The system held no record of what Goozo's position on a notice period was, could not say which template version reflected the current view, and had no way to check an incoming contract against either. Maintaining that layer is work that belongs to a legal function Goozo does not have, so the repository kept storing accurately while the standard behind it went unmaintained.

"Our old system was a mess. It held all of our contracts but sorting them, drafting and tracking was a nightmare. It couldn't tell me what our standard notice period for cancelling an auto-renewal was supposed to be, because nothing in it held that answer. Asking something like 'which of our contracts don't have a DPA' meant looking through multiple possible places and then opening files one at a time. Whatever replaced it had to be something I could run myself, without a lawyer sitting next to me."

Arsaman Bahrami, CEO of Goozo

Goozo's contracts were sound documents. The gap sat above them, in a layer that would hold the company's positions in one place and apply them the same way each time a new agreement came through.

Getting the standard out of the documents Goozo already had

Goozo exported the templates and signed agreements held in the old CLM and uploaded them to Lexnus. Lexnus segmented that language into individual clauses and proposed a category for each one, which gave Goozo a clause library made of the wording it was already using instead of a set of generic samples. Arsaman reviewed and approved the clauses, then built the first playbook from them.

"I didn't need to know what a 'playbook' was going in. I uploaded what we had and reviewed what came back. Two of the templates sitting in the old system carried different auto-renewal notice periods, which is what happens to a template set nobody has owned for a few years, and I had to decide which one was right going forward. That part took real thought. Most of the review was simpler. It was confirming things we had been doing consistently without ever writing them down."

Arsaman Bahrami, CEO of Goozo

The order matters for a company without a legal function. Goozo's positions were taken out of the documents it already had, reviewed once, and approved. Everything Lexnus does after that point runs against that approved standard, whether the contract is one Goozo drafts or one a customer sends back.

How Lexnus is used today

Everything is in one place and he can ask it directly. Lexnus has an assistant mounted on every page of the app, and it answers from Goozo's own clauses, playbooks, contracts, and analysis results. Arsaman can search the wording across the contracts he has imported, pull up what a specific agreement scored and which rules it broke, or ask why something was flagged, and the answer comes back grounded in his own data. The questions that used to mean opening files one at a time are now typed in plain language.

Signed contracts are checked against the current standard. Agreements already in the archive can be run through the same playbook that governs new ones, which shows where an older contract sits outside the position Goozo holds today, including auto-renewal notice periods that predate the current rule. A contract with no DPA comes back flagged as a gap.

New contracts are assembled from the approved clause set. Arsaman answers the questions Lexnus asks about the deal and the document comes back built from clauses that have already been through review. There is no separate legal review step to route it through, because the policy is already inside what gets assembled.

Counterparty paper is checked automatically. When a customer sends back their own version instead of Goozo's, Lexnus runs it against the same playbook and shows what differs from the approved position and why it matters.

"Setting up playbooks is a breeze. I'm not an expert on legal language, and understanding the playbooks seemed overwhelming at first. But with the AI bot I was able to ask it questions, get clarifications, and ultimately ask it to improve whatever needed too."

Arsaman Bahrami, CEO of Goozo

Where it doesn't replace judgment

Lexnus runs the standard Goozo sets. It does not decide what that standard should be. The conflicting auto-renewal notice periods that surfaced during setup still needed Arsaman to choose which one was right going forward. Lexnus made that choice visible and cheap to make once, and the decision was his. Once it was made, the system held it in place for every contract after.

The results

Goozo now holds its contracts and its contract standard in the same system, and the standard is what the system applies. Questions that used to mean opening old files, such as whether a given agreement has a DPA, are answered by running that agreement through the playbook. New contracts come out of an approved clause set. None of it requires anyone at Goozo to know contract law to operate.

"The standard we work to sits in one place now, and everything that comes in gets held to it. I'm not a lawyer, and I was never going to build that layer myself out of Word documents. This is the first time our contracts have felt like something we actively run."

Arsaman Bahrami, CEO of Goozo

The same model at a larger company

Goozo runs this without a legal team. A GC or legal ops function has the same problem from the other side, already knowing the standard but struggling to keep it applied consistently across a larger volume of contracts, more lawyers touching them, and counterparties producing their own paper. The mechanism that helped Goozo is the one that scales there, with more structure around it.

A legal team of several lawyers can run a separate playbook for each contract type it handles, each governed independently and each requiring Admin approval before a change goes live. Clause publishing follows the same approval path Goozo ran informally by having a single decision-maker. At a larger company a Creator drafts or imports a clause and an Admin, usually the GC, approves it before it can appear in any contract. The mechanism holds at any team size. What grows is the number of people accountable to the same fixed standard.

Volume is the other difference. Goozo can fall back on reading a contract by hand when it has to. A company closing dozens a month has no such fallback, which is where deterministic enforcement earns its place. Every contract gets checked against the same published playbook version regardless of who drafted it or which AI tool produced it, and every analysis result stays linked to the exact policy version it was evaluated against. A GC who needs to answer whether a contract was checked six months ago, and against what standard, has that in the audit trail.