Typed Dependency Graphs for Ottawa Real Estate

What You're Looking At

This system models real estate workflows as typed dependency graphs — every stage, service, or milestone is a node, and every prerequisite relationship is a directed edge. A graph engine computes phase ordering, impact propagation, risk analysis, and health status automatically from the structure.

Seven scenarios demonstrate the approach:

1. Transaction Pipeline — a 16-stage dependency graph modelling a seller-side residential resale from listing agreement to closing
2. Agent Operations — a 15-service topology graph modelling the business infrastructure an Ottawa real estate agent depends on
3. Referral Network — a 15-contact dependency graph modelling the agent's professional contacts and referral chains
4. Listing Launch — a 13-task checklist graph modelling the pre-listing preparation workflow from property assessment to go-live
5. Compliance Tracker — a 12-requirement dependency graph modelling RECO regulatory obligations and continuing education
6. Agent Onboarding — a 13-step pipeline modelling the process from pre-registration education to first open house
7. Client Pipeline — a 13-stage funnel modelling the buyer journey from lead capture through closing to referral generation

Same engine. Same patterns. Seven fundamentally different shapes of work.

How This Actually Works (No Math Required)

Imagine you're explaining a real estate deal to a brand new assistant. You'd say things like:

• "You can't list the house until the listing agreement is signed."
• "Photography and the CMA can happen at the same time, but both need the prep to be done first."
• "The buyer's inspection, financing, and lawyer review all happen after the offer is accepted, and all three have to clear before the deal goes firm."

You already think this way. Every experienced agent carries a mental map of what depends on what.

What this system does is write that map down — once — and then ask it questions.

The Map

Each stage in a deal (or each tool in your business) becomes a dot on a map. You draw arrows between them: "this can't start until that's done." That's it. That's the whole input.

You don't tell the system which phase something is in. You don't manually number the steps. You just say "photography depends on property prep" and "MLS listing depends on photography and pricing." The system figures out the rest.

The Questions

Once the map exists, the system can answer questions you'd normally have to think through manually:

• "What happens if financing falls through?" — The system traces forward from "financing-approval" and tells you exactly which stages are affected (conditions-waived, closing) and which are safe to continue (inspection, lawyer review). You already know this, but now it's written down and shareable.
• "What's the single most important thing in this deal?" — The system counts how many things depend on each stage. Listing agreement: 15 things depend on it. Photography: 11. Financing: only 2. Now you know where to focus your attention.
• "MLS went down — what can I still do?" — The system traces the operations map and tells you: CRM contacts, email marketing, trust accounting, e-signing — all still work. Website, transaction management, and lead tracking — broken until MLS comes back.
• "Where's the bottleneck?" — Any stage with no alternative path around it is a single point of failure. In a deal, that's almost everything (deals are sequential). In your business operations, it's OREB membership — the one thing 10 other services depend on with no backup.

Why Bother?

Three reasons:

1. Shared understanding. When you hand a deal to a colleague, or explain your business dependencies to your brokerage, the map is the explanation. No ambiguity about what depends on what.
2. Automatic answers. Change the map (add a stage, remove a dependency) and every analysis updates automatically. There's no spreadsheet formula to fix, no report to regenerate manually.
3. The questions you don't think to ask. Most agents know their biggest risks intuitively. But "photography vendor is a single point of failure that blocks 11 downstream stages" — that's the kind of thing that only becomes visible when you map it explicitly.

Scenario 1: Transaction Pipeline

The Deal

A seller-side residential resale in Ottawa — from signed listing agreement through to closing. 16 stages across 11 transaction phases, with dependencies computed (not manually assigned) from the graph structure.

Cost and Risk Data

Each stage now includes real Ottawa cost estimates and failure rate percentages:

Ottawa Market Context

• Board: OREB (Ottawa Real Estate Board)
• MLS System: Matrix
• Regulatory: RECO (Real Estate Council of Ontario), REBBA 2002
• Standard Forms: OREA 200 (Listing), 100 (Offer), 101 (Amendment), Schedule A (Conditions), 220 (SPIS)
• Typical Timelines: 90-day listing, 5-day inspection condition, 10-day financing condition, 60-day closing

What This Answers

"Which stages delay closing the most if they slip?"

The graph computes downstream impact for every node:

The listing agreement is the ultimate dependency — everything cascades from it. Market analysis and property prep are the next-highest leverage items because the entire marketing and showing sequence depends on both being complete.

"The buyer's financing fell through — what's the blast radius?"

• 2 stages directly affected: conditions-waived, closing
• Safe to continue: home-inspection, lawyer-review (they're peer conditions, independent of financing)
• Recovery sequence: buyer finds new lender or deal collapses and listing goes back to active

"Inspector found foundation cracks — deal may not survive"

Health propagation through the graph: home-inspection marked as down, everything upstream stays healthy, everything downstream (conditions-waived, closing) propagates to at_risk. The buyer can walk away, renegotiate, request repairs, or waive.

"Which stages are single points of failure?"

Because the transaction pipeline is largely linear (each stage depends on the previous), most stages are structural SPOFs — there's no parallel path around them. The graph identifies 10 SPOFs, with the highest-leverage ones being:

Sample Listings

The transaction model includes real Ottawa inventory from Royal LePage Team Realty to ground the analysis in concrete data:

Scenario 2: Agent Operations

The Platform

15 services across 6 deployment layers, modelling the complete technology and compliance infrastructure an Ottawa real estate agent depends on:

Cost and Alternatives

Each service includes monthly cost estimates and backup options:

Regulatory Context

• Licensing: RECO (Real Estate Council of Ontario) — 12 MCE credits per 2-year cycle
• Governing Act: REBBA 2002 (Real Estate and Business Brokers Act)
• Insurance: Errors & Omissions — mandatory through RECO
• Trust Accounting: Brokerage-managed, deposits per REBBA
• Document Retention: 6 years (RECO requirement)
• Compliance: FINTRAC (cash transactions), PIPEDA (privacy), CASL (anti-spam)

What This Answers

"Matrix MLS is down — what can't I do?"

• Multiple services affected: website (IDX feed dies), photography-vendor (can't upload), transaction-mgmt (can't pull listing data), and everything downstream
• Safe to continue: CRM contacts, email marketing, trust accounting, e-signing — these don't depend on live MLS
• Recovery order: MLS back online first, then transaction management and website resume, then marketing catches up

MLS is the backbone of daily operations. When it goes down (OREB reports scheduled maintenance windows), the blast radius is significant but not total — the graph shows exactly what's safe to keep working on.

"My CRM crashed — what's broken?"

• E-signing, website, email-marketing, social-media, lead-tracking, and document-archive are all downstream of the CRM
• MLS access, trust accounting, and photography vendors are independent — you can still list and show properties

"Which services are single points of failure?"

Scenario 3: Referral Network

The Network

15 professional contacts modelling the agent's referral relationships — mortgage brokers, lawyer, inspectors, stager, photographer, appraiser, insurance agent, contractor, title insurance, moving company, surveyor, cleaning service, and social media marketing. Dependencies represent referral chains and workflow ordering.

What This Answers

"My mortgage broker is on vacation — what stalls?"

• 2 contacts affected: Lisa Tran (appraiser, loses referral pipeline) and Kate Brown (insurance, loses coordination)
• Safe to continue: James Wu can handle mortgage pre-approvals directly, plus all legal, inspection, and marketing contacts are independent
• Key insight: James Wu is a backup for mortgage work itself, but not for the referral pipeline Sarah feeds

"What if my lawyer retires?"

• 3 contacts affected: title-plus (Stewart Title), ottawa-moving, mark-reid (surveyor) — all routed through Mike
• No backup lawyer — unlike mortgage and inspection, there's no second option established
• Mitigation: Establish relationship with a second real estate lawyer immediately

"Tom is booked solid — what's the cascade?"

Tom Zhang anchors the longest chain in the network (depth 3):

• Tom (stager) → Amy Wilson (photographer) → Social Boost (social media)
• Tom → Clean Sweep (cleaning coordination)
• No backup stager — the entire marketing pipeline stalls. Consider virtual staging as a stopgap.

"Who are my most critical contacts?"

Response Times and Capacity

Each contact now includes operational data — how fast they respond and how much work they can handle:

Scenario 4: Listing Launch

The Checklist

13 tasks across 7 phases modelling the pre-listing preparation workflow — from initial property assessment through staging, photography, and MLS entry to social media launch:

What This Answers

"The photographer cancelled — what's delayed?"

• 3 tasks blocked: MLS entry, social campaign, and signage all wait on photos
• Virtual tour is a peer (same phase) — also stalled if it depends on the same staging
• Mitigation: Have a backup photographer on standby during busy season

"Tom is booked and can't stage this week — what cascades?"

• 5 downstream tasks stall: photography, virtual tour, MLS entry, social campaign, signage
• Everything upstream is safe: CMA, declutter, repairs, floor plans, deep clean can all proceed
• Staging is the biggest bottleneck in the launch checklist — it gates the entire marketing phase

Scenario 5: Compliance & Education

The Requirements

12 regulatory requirements across 4 layers modelling the compliance obligations an Ontario real estate agent must maintain:

What This Answers

"My RECO registration lapses — what cascades?"

• 10 requirements affected — everything downstream of registration falls away
• You literally cannot practice — E&O insurance, OREB membership, brokerage agreement all require active registration
• Recovery: Re-register with RECO, then re-establish each downstream requirement in dependency order

"I'm behind on MCE credits — what's at risk?"

• Annual declaration blocked — can't complete without credits + active E&O
• Registration renewal at risk — RECO won't renew without completed MCE
• Cost to catch up: $500–$1,000 per 2-year cycle, available from OREA and approved providers

Scenario 6: New Agent Onboarding

The Pipeline

13 steps across 8 phases modelling the full onboarding journey for a new real estate agent joining Royal LePage Team Realty:

What This Answers

"RECO is slow processing the application — what's blocked?"

• 10 steps stall — everything from brokerage interview forward waits on RECO approval
• Nothing can proceed — RECO approval is the external gate that controls the entire timeline
• Typical delay: 2–4 weeks, but can stretch to 6+ weeks during peak registration periods (January, September)

"Brokerage paperwork is delayed — what's bottlenecked?"

• 8 steps blocked: OREB registration, CRM, email, headshot, mentor, MLS training, website, first open house
• Brokerage agreement is the widest fan-out node — more downstream dependencies than any other step

Scenario 7: Client Pipeline

The Journey

13 stages across 11 phases modelling the full buyer client lifecycle — from lead capture through qualification, showings, and closing to post-sale referral generation:

What This Answers

"My mortgage broker is backed up — what stalls in the client pipeline?"

• 7 downstream stages blocked — property search, showings, offer, contract, closing, nurture, referral
• Pre-approval is the critical gate — without it, you can't seriously search for properties
• Mitigation: Use James Wu (CIBC backup) for overflow

"Where do I lose the most clients?"

The conversion rates tell the story. The biggest drops:

• Offer acceptance: 50% — in a competitive Ottawa market, half of offers don't succeed
• Online leads to consultation: 2–5% — most online leads never convert (sphere referrals are 7× more effective)
• Showings to offer: 60% — buyer fatigue, changing requirements, market shifts

Seven Shapes, One Engine

97 total nodes, 104 edges, 7 scenarios — all from the same pattern library.

What This Offers That Existing Tools Don't

This system doesn't replace Dotloop, kvCORE, or Matrix MLS. It does something different — it models the relationships between stages and services, not just the stages themselves.

The key insight: agents already know intuitively that "if financing falls through, we lose the deal." What they don't have is a tool that quantifies it — shows exactly which 2 downstream stages are affected, which 2 peer conditions are safe, and what the recovery sequence looks like. The graph makes implicit knowledge explicit and queryable.

The Interactive Explorer

Open the Explorer →

The interactive graph explorer renders all seven scenarios as layered or force-directed network diagrams:

• Seven tabs: Transaction, Operations, Referral, Listing, Compliance, Onboarding, Client
• Node selection: click any node to see its details — description, types, owner, dependencies, dependents, cost, and enrichment data
• Layer filtering: toolbar buttons toggle visibility by phase/layer
• Layout toggle: switch between layered (phases top-to-bottom) and force-directed (organic clustering by layer)
• Deep linking: share a specific node — e.g., #transaction/financing-approval, #listing/staging, #compliance/reco-registration, #client/showings
• Mobile-friendly: bottom sheet interface on phones, side panel on desktop

All graph analytics are pre-computed and served as static JSON. The explorer is a pure renderer — no server, no database, no login required.

Technical Notes

What This Is Built On

The system uses CUE, a typed configuration language originally developed at Google for infrastructure management. The same graph patterns that model datacenter dependency networks are applied here to model real estate transaction workflows and agent business infrastructure.

The patterns are reusable: the graph engine computes depth, ancestors, and topology; the deployment planner generates phased execution sequences; the impact query traces failure propagation; and the SPOF detector finds nodes with no redundancy. These patterns work identically whether the nodes are transaction stages, business services, work packages, or any typed resource with dependencies.

Running Queries Locally

With CUE v0.15.3 installed: