Build or buy legal AI? The honest test for an in-house team

Almost every in-house legal team is now deciding this, not whether to use AI at all. The market has two shapes on offer — buy a platform your lawyers log into, or build the capability into the systems they already work in — and the honest answer is that both are right, for different teams. Here is the test that tells you which one is yours, written by people who build the second kind and will tell you plainly when the first is the better call.

When buying a platform is the right answer

A horizontal legal-AI platform is the faster, cheaper, lower-risk choice when your work is document-first and system-light. Buy when:

• Your lawyers' day is drafting, research and review, and the output is a document — not a record that has to update something else.
• You want value in weeks with no engineering, and a vendor who maintains the model, the updates and the security posture for you.
• Your matters do not need to read from or write back to your ERP, your finance system or your CRM — the contract is the deliverable, not a trigger for a downstream process.

If that is your team, buy the platform. It is a solved problem, the good ones are very good, and building your own would be spending engineering to reinvent it. We will say so in the first conversation.

When building into your ERP wins

The calculus flips when the work is system-first — when the contract is not the end of the job but the start of one that lives in your operational systems. Build into your ERP when:

• Your contracts and obligations already live in Dynamics 365, SAP or Business Central, attached to purchase orders, vendors and finance records — and the value is in keeping the record and the agreement in step, not in producing another document.
• A renewal date needs to become a field on the record, an obligation needs to sit against the vendor it binds, and a price on a PO needs to be checked against the price in the signed paper. A document-first platform cannot reach into the system to do that.
• Your team will not adopt a second place to work. The most expensive legal AI is the one nobody opens — a tool outside the existing workflow gets used for a week and abandoned. Building into the system they already live in removes the switch entirely.
• The agreements cannot leave your environment — regulatory, privilege or policy — so the capability has to run inside your own tenancy, on your own data.

That is not a horizontal platform's job, and it is not a criticism of one. A product sold to thousands of firms cannot ship pre-integrated into your particular Dynamics or SAP estate, your approval routing, your paper. That integration is an engineering job — and it is the one we do.

The test in one line

Is the contract the deliverable, or the trigger for work in another system?

If it is the deliverable, buy a platform. If it is the trigger — if the value is in your ERP staying true to your agreements — build it in, on your own data, where your team already works.

What we build, when the answer is "build"

The capability is the pipeline behind AI contract review, pointed at your own systems: clauses matched against your playbook by meaning rather than clause number, obligations and renewal dates extracted as structured data with the source clause behind each one, a person committing every change, all running inside Dynamics 365, SAP or Business Central rather than beside it. The engineering is written up in full in how legal AI is engineered.

We built one to prove it ships: Paralegent AI reads every clause against a playbook and drafts the replacement language, in production today — evidence, not a product you have to buy. What we build for you is that pipeline on your paper and your positions.

If your contracts live in your ERP and you are weighing this, tell us what you run and where you are stuck — you will get a straight answer, including "buy the platform" when that is the right one. Or see Legal AI Ops for Dynamics 365 and SAP.