Challenge
Chipotle needed to seed its existing database with structured, per-site data drawn from a legacy portfolio of decades-old, non-standard scanned leases from hundreds of landlords. For every store, across 31 service categories, it needed to know who executes the service, who pays, and where the lease says so—information that lived only inside individual PDFs.
Approach
The full portfolio was ingested into one Legora workspace and OCR'd end to end. Axiom worked category-first, not lease-first—building 31 parallel Legora tabular reviews, each running against all 1,945 stores in a single pass. Every category passed a three-stage attorney-led QC cycle: a 10-site sample, a validation set to catch drift on non-standard forms, then the full run, normalized to Chipotle's taxonomy.
Impact
-
Delivered in 20 weeks —work Chipotle estimated at 18 months for one paralegal.
-
Obligations moved from 1,945 individual documents to a single filterable master file.
-
All 31 outputs consolidated into one workbook, keyed by store and service type.
- Engineered to load directly into Chipotle's new database.
Every question about our obligations under our legacy leases required that a team member individually review the documents and provide an analysis. Now every obligation sits in one place, and we can answer a portfolio-wide question in minutes.
- Betsy Waterman, Real Estate Assistant General Counsel, Chipotle
*Client estimate: ~78 weeks (18 months) for one paralegal. Delivered in 20 weeks (5 months) with a two-person AI-enabled Axiom team.
