Facilities Platform
What tools does the Head of Facilities need to level the asymmetry?
Three Layers
The facilities function runs on three stacked layers. Each changes the buyer-side leverage available.
- System of record — the lease ledger, the property register, the incident log, the vendor contract list, the building documentation library. Source of operational truth.
- Agent layer — AI workers that read leases, run NPV models, benchmark rents, sweep public records on counterparties, and prep negotiation playbooks. The leverage that collapses years of professional advantage into hours of prompt work.
- On-chain instruments — emerging crypto rails: title registries, lease registries, smart-contract escrow, stablecoin settlement, DeFi lending against lease value. Sell-side / tokenization is mature. Buyer-side is nascent.
AI Tools That Work Today
Highest-ROI single move first.
Lease clause analysis
- What it does: LLM reads any lease, flags non-standard / tenant-unfriendly clauses, benchmarks against standard tenant-side redlines.
- Why it matters: Catches 10-year cost commitments hidden in single sentences. Two hours of prompt work routinely catches $50k+ exposures.
- Who uses it: Head of Facilities + property lawyer (the lawyer redlines from a richer brief).
- Maturity: Works today on any commercial-grade LLM.
Total-cost-of-occupancy NPV builder
- What it does: LLM builds the NPV model per the Process Lens A. Operator fills inputs. Sensitivity runs in seconds.
- Why it matters: Replaces a $5k accountant engagement with 30 minutes. Forces actual quantification of TCO before committing.
- Who uses it: Head of Facilities + CFO.
- Maturity: Works today.
Counterparty due diligence
- What it does: LLM-driven public-record sweep on landlord entity — reviews, news, court filings, previous tenant complaints, regulatory actions.
- Why it matters: Surfaces landlord risk that the broker will not surface. Would cost $2k+ from a PI; LLM does it in an afternoon for query cost.
- Who uses it: Head of Facilities + property lawyer.
- Maturity: Works today.
Negotiation prep playbook
- What it does: LLM brief on landlord's recent deals, vacancy pressure, motivations. Generates counter-arguments and walkaway anchors for every clause the landlord will propose.
- Why it matters: The asymmetric counterparty walks in with a playbook. The amateur tenant walks in with hope. AI closes that specific gap.
- Who uses it: Head of Facilities + tenant rep broker + property lawyer.
- Maturity: Works today.
Comparable-rent synthesis
- What it does: Aggregates published comps, tenant-reported actuals, public lease filings (where available) + LLM synthesis to estimate real market rent.
- Why it matters: Asking rent and actually-paid rent differ. The gap is the landlord's negotiation cushion.
- Who uses it: Head of Facilities + tenant rep broker.
- Maturity: Closes some of the gap today. Jurisdiction-dependent. Compstak / CoStar and local equivalents help.
Workplace density audit
- What it does: Photo of current space + headcount → AI-driven m² recommendation. Often 20–40% less than broker estimates suggest.
- Why it matters: The broker is incentivised toward more m². The density audit is incentivised toward right-sized cost.
- Who uses it: Head of Facilities + workplace strategist (when external).
- Maturity: Growing.
Fit-out cost reality check
- What it does: AI takeoff from photos / drawings + AI cost estimate, compared against contractor quotes.
- Why it matters: Catches the inflated quote and the missing scope, both common.
- Who uses it: Head of Facilities + QS.
- Maturity: Growing.
Crypto Rails — Emerging Instruments
Track, do not bet. The investor / sell-side has tokenized fractional ownership and RWA tokens. The operator / buyer-side is mostly nascent.
| Asymmetry today | Crypto primitive | Buyer / tenant benefit | Status |
|---|---|---|---|
| Title fraud / hidden encumbrances | On-chain title registry | Single source of truth — bypass title insurance fees | Nascent — see Real Estate Industry Friction Map |
| Comparable rent opacity | On-chain commercial lease registry | Real rents, real terms, real incentives — not asking rents | Does not exist yet — open lane |
| Cross-border purchase friction | Stablecoin settlement | Settle in days not months | Working in EU/SG, blocked in US |
| Lease enforcement asymmetry | Smart contract escrow | Rent-free, fit-out contribution, break-clause triggers — automated not discretionary | Conceptual |
| Tenant reputation portability | On-chain payment history | Portable tenant credit score across landlords and jurisdictions | Does not exist yet |
| Landlord reputation portability | On-chain landlord rating | Tenants leave verifiable reviews; bad actors lose tenants | Does not exist yet |
| Capital access for fit-out | DeFi lending against lease | Borrow against lease value, not personal guarantee | Early |
What to Skip
Tokenized fractional CRE for the operating business. For an operator buying or leasing their own premises, the unit of value is the lease term, not the equity slice. Tokenization solves investor fractionalization problems, not operator control problems. A chef with 1% of their shopfront via a token has worse control than a chef on a strong 10-year lease.
The Real Estate Tokenization canon covers the sell-side / investor-side platforms (DigiShares, Lofty, Homebase, Manifest, Parcl, reAlpha, Sky Trade, Flow Life). None of them solve the buyer-side asymmetry. That's the open lane this Platform pillar tracks.
Stack by Operating Stage
The platform layer thickens as the business scales. Start lean.
- Seed (1–10 FTE): LLM + spreadsheet. Single-location lease decisions run on AI lease analysis + AI NPV builder + counterparty due diligence + negotiation prep. No dedicated facilities team yet — founder + lawyer + tenant rep.
- Growth (10–50 FTE): Add a lease calendar tool, an incident log, vendor management. Bring a workplace strategist in for density audits at each renewal. AI tools above are still doing the heavy lifting.
- Scale (50+ FTE): Dedicated Head of Facilities. Lease portfolio in a real system of record. Recurring density / TCO reviews. Counterparty database institutionalised. AI agents embedded in the workflow rather than ad-hoc.
The stack does not justify itself until the gauge it serves is firing. Add platform tools when a Performance gauge starts breaking — not before.