Legal document review · Production

AMD uses Luminance for AI-powered contract mark-up, risk identification, and clause search across 2,400+ NDAs annually

The problem

AMD's small legal team handles approximately 85% of the company's agreements, including roughly 2,500 NDAs and 300–400 commercial agreements per year, plus software license reviews from 25,000 engineers — a volume too high to manage without tooling.

Workflow diagram · grounded in source
1
Third-party NDA arrives
Trigger
Any third-party NDA that comes in goes through Luminance.
▾ source quote
“any third-party NDA that comes in, it just goes through Luminance”
2
AI first mark-up
Ai action
Luminance performs the first mark-up to handle commodity-based contract review work.
▾ source quote
“Let Luminance do the first mark-up to take some of the work that is commodity-based work”
3
AI risk identification
Ai action
Luminance identifies the known risk issues present in the contract.
▾ source quote
“Let the AI identify the issues that we know are there all the time”
4
Lawyer high-value negotiation
Human review
Lawyers and paralegals handle high-value negotiation after the AI completes the first mark-up.
▾ source quote
“let the lawyer or the paralegal do the high value negotiation”
5
Risk-based routing (planned)
Routing
A planned workflow would route software licenses through Luminance with green-, yellow-, or red-light decisions based on identified risks.
▾ source quote
“Wouldn't it be great if we just take those licenses, run them through Luminance and just say green light, you can proceed. Any identified risks and it has to go to the legal department. Or yellow light might mean it …”
6
Natural language clause search
Ai action
Lawyers search past contract clauses using natural language queries directed at Lumi.
▾ source quote
“hey Lumi, pull up all my feedback clauses that I have written in the last five years”
Reported outcome

AMD now routes all third-party NDAs through Luminance for first mark-up and risk identification, with lawyers focusing on high-value negotiation; contract clause searches that previously took half an hour are now handled instantly via Lumi.

Reported metrics
share of AMD agreements through legal team85%
annual NDA volume2,500
annual NDA volume (second figure)2,400
Annual commercial agreements3 or 400
Show all 8 reported metrics
share of AMD agreements through legal team85%
annual NDA volume2,500
annual NDA volume (second figure)2,400
annual commercial agreements3 or 400
engineers sending software licenses for legal review25,000
risk identification quality vs. paralegalas good a job as my paralegal can at identifying the risks
competitive advantage over evaluated alternativesfar, by far above the competition
time previously spent searching for past contract clauseshalf an hour looking
Reported stack
LuminanceLumi
◆ Does this fit your context?

Compare to your context

Tell us your scale, team, and constraints. We'll show what changes at your size, what fails at your scale, and whether this case is a fit, needs adaptation, or won't scale to you. Free demo, no signup.

Compare to your context →
~30 seconds · free
Source
https://www.luminance.com/customers/luminance-is-far-above-the-competition/
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

AMD now routes all third-party NDAs through Luminance for first mark-up and risk identification, with lawyers focusing on high-value negotiation; contract clause searches that previously took half an hour are now hand…

What tools did this team use?

Luminance, Lumi.

What results were reported?

share of AMD agreements through legal team: 85%; annual NDA volume: 2,500; annual NDA volume (second figure): 2,400; Annual commercial agreements: 3 or 400 (source-reported, not independently verified).

How is this legal document review AI workflow structured?

Third-party NDA arrives → AI first mark-up → AI risk identification → Lawyer high-value negotiation → Risk-based routing (planned) → Natural language clause search.

Related legal document review cases

WHAT TO DO WITH THIS

Now compare it to your context

This case is one data point. Whether its pattern fits you depends on your volumes, your stack, and your exception load — that comparison is the step no case study can do for you.

Browse the Legal document review cases →