Skip to main content

Case Studies

See how mixus Legal AI handles common legal document scenarios. These examples demonstrate our capabilities across different document types and modification requests.

Case 1: NDA Modification

Scenario

A startup needs to modify their standard NDA to make confidentiality obligations mutual instead of one-way, and extend the confidentiality period from 2 years to 5 years.

Request

“Make this NDA mutual. Both parties should have the same confidentiality obligations. Also extend the confidentiality period to 5 years after termination.”

What Happened

1

Classification

Document identified as NDA with 95% confidence. Detected one-way structure with Discloser/Recipient terminology.
2

Changes Applied

  • Updated definitions to be party-neutral - Modified all confidentiality obligations to be mutual - Changed “Discloser” and “Recipient” to “each party” - Extended confidentiality period from 2 to 5 years - Updated signature blocks for mutual execution
3

Validation

All 5 passes completed successfully. No unexpected changes. One medium-risk flag for extended confidentiality period (noted but approved).
4

Delivery

Three files delivered: redline DOCX (12 changes marked), clean DOCX, and blackline PDF.

Change Summary

SectionChanges
Definitions3 terms updated for mutuality
Confidential Information2 modifications for mutual scope
Obligations4 changes for mutual obligations
Term1 change (2 years → 5 years)
Miscellaneous2 updates to signature blocks

Case 2: MSA Section-Specific Edit

Scenario

A legal team needs to modify only the liability section of an MSA to add a carve-out for gross negligence and willful misconduct.

Request

“In Section 9 (Limitation of Liability) only, add a carve-out stating that the liability cap does not apply to damages arising from gross negligence or willful misconduct.”

What Happened

1

Classification

Document identified as MSA with 92% confidence. Section 9 located and identified as Limitation of Liability provision.
2

Section Targeting

Edit restricted to Section 9 only. All other sections (including related indemnification in Section 8) remained untouched.
3

Changes Applied

New subsection 9.4 added: “Notwithstanding the foregoing, the limitations set forth in this Section 9 shall not apply to damages arising from a party’s gross negligence or willful misconduct.”
4

Validation

Section compliance verified—no changes outside Section 9. Risk analysis flagged as medium risk (liability cap exception added).

Section Compliance Report

Sections Specified: Section 9
Sections Modified: Section 9 only
Out-of-Bounds Changes: None
Compliance Status: Compliant ✅

Case 3: Employment Agreement Revision

Scenario

HR needs to update an employment agreement to change the non-compete geographic scope and add a new intellectual property assignment provision.

Request

“1. In the non-compete section, change the geographic restriction from 50-mile radius to the states where the company has offices (California, Texas, and New York).
  1. Add a standard intellectual property assignment clause that assigns all work product to the company.”

What Happened

1

Classification

Document identified as Employment Agreement with 88% confidence. Located existing non-compete (Section 7) but no IP assignment section.
2

Non-Compete Edit

Section 7.1 modified: “50-mile radius of any Company office” changed to “the states of California, Texas, and New York”
3

IP Assignment Addition

New Section 8 (Intellectual Property) added with standard assignment language. Subsequent sections renumbered accordingly.
4

Risk Analysis

  • Non-compete change: Low risk (narrower geographic scope) - IP assignment: Medium risk (new obligation for employee) Both changes flagged for review by senior HR.

Risk Report Excerpt

## Medium Risk: Intellectual Property Assignment

**Type:** New obligation added
**Section:** Section 8 (new)

Standard IP assignment clause added assigning all work
product, inventions, and developments to the Company.

**Recommendation:** Review with employment counsel to ensure
compliance with state-specific IP laws, particularly California
Labor Code Section 2870.

Case 4: Document Comparison

Scenario

Outside counsel sent back a revised contract. The legal team needs to see exactly what changed between the original version and the counsel’s markup.

Request

“Compare our original MSA with the version counsel returned. Show me all differences.”

What Happened

1

Document Loading

Both documents loaded: original (Acme-MSA-v1.docx) and revised (Acme-MSA-counsel-v2.docx).
2

Comparison

Deterministic comparison performed—no AI interpretation, just direct document comparison showing all differences.
3

Output

Redlined document showing 23 changes across 8 sections. Change statistics: 15 additions, 6 deletions, 2 modifications.

Comparison Statistics

MetricCount
Total Changes23
Additions15
Deletions6
Modifications2
Sections Changed8 of 14

Case 5: Complex Multi-Section Edit

Scenario

A company is pivoting their business model and needs comprehensive updates to their SaaS agreement to reflect new pricing, service levels, and data handling provisions.

Request

“Update the agreement to reflect our new pricing model:
  • Change from per-seat to usage-based pricing
  • Update the SLA to 99.9% uptime (was 99.5%)
  • Add GDPR data processing addendum references
  • Modify termination to allow 30-day notice (was 90 days)“

What Happened

1

Classification

Document identified as SaaS Agreement with 91% confidence. All target sections located.
2

Multi-Section Changes

  • Section 4 (Fees): Pricing structure rewritten for usage-based model - Section 6 (SLA): Uptime commitment updated to 99.9% - Section 7 (Data): New GDPR DPA reference added - Section 11 (Term): Notice period changed to 30 days
3

Validation

Required 2 retry attempts due to scope creep in initial attempt (validation caught changes bleeding into Section 8). Final version passed all checks.
4

Risk Analysis

  • SLA improvement: Low risk (more favorable to customer) - Shorter notice period: High risk (reduces planning window) High-risk item flagged for executive review.

Validation History

Attempt 1: Failed (section compliance - changes in Section 8)
Attempt 2: Passed all checks
Total Retries: 1
Final Status: Success

Key Takeaways

Section Targeting Works

When you specify sections, changes stay within those boundaries.

Risk Detection is Automatic

High-risk changes are always flagged for attention.

Validation Catches Issues

Our retry mechanism catches and corrects problems before delivery.

Comparison is Deterministic

Document comparison shows exactly what changed—no AI interpretation.

Try It Yourself

Upload a document and see how mixus Legal AI handles your use case.