# How We Deconstruct a Lease to Reconstruct It Better Blog | LeasePilot [Blog](/blog)Technology # How We Deconstruct a Lease to Reconstruct It Better A technical look at LeasePilot's layered document architecture and why separating template, building, state, and deal data enables unprecedented flexibility. ![Lior Kedmi](/_next/image?url=%2Fleadership%2Flior-kedmi.jpg&w=3840&q=75&dpl=dpl_2uqrzvFtdfjJy2rKqbPgTzBd5aYQ) Lior Kedmi CTO September 16, 20248 min readCopy link TL;DR A commercial lease isn't one document, it's five layers mixed together. We decompose them, store them separately, and reconstruct on demand. This architecture is why we can handle complexity that breaks other systems. § 01 ## [The Monolithic Document Problem](#the-monolithic-document-problem) Open any commercial lease in Word. What do you see? A single document. Pages of text. Clauses flowing one after another. But that's an illusion. What you're actually looking at is multiple distinct things fused together: - **Template structure**: Your standard lease form, the clauses you've refined over years - **Building information**: This specific property's address, square footage, common areas - **State-specific language**: California disclosures that don't apply in Texas - **Deal terms**: This tenant's rent, term length, options, concessions - **Calculated values**: Rent schedules, pro-rata shares, escalation tables In a Word document, these are all mixed together. Change your form? You have to update every variant manually. Acquire properties in a new state? Hope you remember all the compliance requirements. Update a building's common area factor? Good luck finding every lease that references it. This is the architecture of fragility. § 02 ## [The Five Layers](#the-five-layers) LeasePilot's fundamental architectural decision: **decompose the lease into its constituent layers and store them separately.** ### Layer 1: Your Lease Forms Your lease forms represent years of legal refinement. Clause language negotiated through disputes. Provisions added after problems surfaced. Fallback positions documented through experience. In LeasePilot, your forms are stored as structured objects, not Word files. Each clause is a discrete unit with: - The clause text itself - Metadata about when to include it - Variant options for different scenarios - Fallback positions for negotiation When you update a clause, it updates everywhere that clause is used, instantly, consistently. ### Layer 2: Building Information Every property has characteristics that flow into leases: - Address and legal description - Rentable and usable square footage - Common area definitions - Building systems and specifications - Landlord entity information This information lives in one place. When a building's common area factor changes, every lease for that building reflects the change automatically. No hunting through documents. ### Layer 3: State-Specific Requirements Commercial leasing is governed by state law. California requires specific seismic disclosures. New York has unique commercial tenant protections. Florida hurricane provisions differ from Texas. LeasePilot maintains a compliance layer, state-specific requirements that automatically include (or exclude) based on property location. When California updates its disclosure requirements, we update one place. Every California lease reflects the change. ### Layer 4: Deal Terms The business terms of this specific transaction: - Base rent and escalation structure - Lease term and option periods - Tenant improvement allowances - Concessions and free rent periods - Security deposit requirements Deal terms are entered once and flow throughout the document, into the body, the schedules, the exhibits. No re-entering the same number in twelve places. ### Layer 5: Calculated Values Leases contain complex calculations: - Rent escalation schedules over a 10-year term - Pro-rata share calculations - Percentage rent breakpoints - Operating expense estimates These calculations are performed by the system, not typed by hand. The inputs come from other layers (base rent from deal terms, RSF from building information). The outputs populate automatically. This is where the deterministic vs. probabilistic distinction matters most, lease calculations must be exact, every time. § 03 ## [Why Separation Matters](#why-separation-matters) This architecture solves real problems. ### Problem 1: Template Updates **The old way**: You improve a clause. Now you have to update every variant of your form, then hope attorneys use the new version, then audit to make sure old language isn't still floating around. **With layers**: Update the clause once. Every lease generated from that point forward uses the new language. Consistency is automatic. ### Problem 2: Portfolio Acquisitions **The old way**: You acquire 40 properties. Each has leases drafted by someone else, using their forms, their language, their inconsistent approaches. **With layers**: Import the building data. New leases for those buildings use your forms, your clauses, your standards, immediately. ### Problem 3: Multi-State Compliance **The old way**: Your attorney remembers that California requires certain disclosures. Hopefully. And remembers to remove them for Texas properties. Hopefully. **With layers**: State compliance is systematic. California property? California requirements included automatically. No memory required. ### Problem 4: Audit and Analysis **The old way**: "Which leases have the old co-tenancy language?" Someone opens every lease and reads through it. **With layers**: Query the structured data. Which leases used clause version 2.3? Here's the list. Which buildings use the pre-2024 operating expense definitions? Here they are. § 04 ## [A Real Example](#a-real-example) A client updates their CAM reconciliation language after a dispute revealed ambiguity. **In a traditional workflow**: They have 15 form variants across retail, office, and industrial. An attorney manually updates each one. Existing leases in negotiation? Those need to be manually updated too. Leases signed before the change? Those have the old language forever. **In LeasePilot**: They update the CAM clause once. Every form that references that clause now generates with the new language. Leases in draft automatically reflect the change. They can run a report showing which executed leases have the old language for risk tracking. Time to implement: minutes vs. weeks. § 05 ## [The Flexibility This Enables](#the-flexibility-this-enables) Layered architecture isn't just about efficiency. It enables capabilities that aren't possible with monolithic documents. **Complex conditionals**: "If this is a California retail property over 10,000 SF with percentage rent, include these specific provisions." The system evaluates conditions across layers to assemble the right document, the kind of deal logic LeasePilot can encode for your specific portfolio. **Amendment intelligence**: An amendment is a delta, changes to specific layers. The system knows what changed from the original lease because it has structured data, not just two Word documents to compare. **Portfolio analytics**: Because deal terms are structured data, not text buried in documents, you can analyze across the portfolio. Average TI allowance by property type? Rent escalation patterns? Option expiration concentrations? The data exists to answer these questions. § 06 ## [The Technical Trade-off](#the-technical-trade-off) This architecture has a cost: complexity in the build. A Word template is simple, anyone can edit it. A decomposed, layered system requires more engineering to build and maintain. That trade was made intentionally. The complexity lives in the platform so it doesn't live in your workflow. The attorney doesn't need to think about layers, they see a lease. But behind that lease is an architecture that makes consistency, compliance, and flexibility possible at scale. * * * The lease that looks like one document is actually five things interleaved. By separating them, storing them properly, and reconstructing them intelligently, the system handles complexity that Word-based approaches can't touch. Ten years in, this architectural decision remains the foundation of everything LeasePilot does. § Adjacent reading ## More from the ledger [§ 01DEC 18, 2024 Technology ### Supporting Any Complexity in Lease Documents Lior Kedmi9 MIN READ Read →](/blog/supporting-any-complexity-in-lease-documents) [§ 02MAR 06, 2024 Technology ### How LeasePilot Actually Works: A Technical Overview Lior Kedmi10 MIN READ Read →](/blog/how-leasepilot-actually-works-technical-overview) [§ 03FEB 06, 2025 Technology ### Why Generic Document Automation Failed CRE Legal Teams David Saltman7 MIN READ Read →](/blog/why-generic-document-automation-failed-cre) § See it in practice ## Reading about it is one thing. Watching it happen is another. See LeasePilot draft a lease in your team’s own templates, with your clauses and your defaults. [Schedule a Demo](/demo)