You laid out the seven stages on May 14. Here's how we partner across all of them.
Stages 1–2 · core paid engagement | Stages 3–7 · same foundation, optional extensions | The closing · stays human
The closing stays human. The relationship-building is yours. Everything else compounds on the foundation we'd build together. This page is the conversation, in writing, before we have it — every quote receipted, every research file inventoried, every open question listed. June 5 can be substantive instead of introductory.
This is the framework Kareem wrote on May 14, confirmed twice. Stages 1 and 2 are the entire scope of the core engagement. Stages 3 through 7 are how the foundation extends — each links to a deeper page on this site.
Most CRMs go silent after close. Ours stays alive. Acquisition is expensive; retention & expansion at known accounts is where the next decade's deal economics actually sit.
The funnel doesn't end — it loops. Stage 7 is where margins live; Stage 1 is where the next deal starts. See the full closed-loop story →
Three Colliers conversations are documented, ingested, and queryable. Everything below is grounded in those exact words. This page itself was assembled from the transcript & email corpus, not invented in a slide deck.
"We are actively seeking a solution on creating a 'qualified opportunity list' which is informed by the interaction of: CRE tenant signals / triggers with a connection to a relationship (relationships being the foundation of success). Your solution — land meetings."Kareem Dossa, email to Anis, 5/14/2026
"I'm only looking for that top of the funnel business intelligence. I don't need all the other stuff. I can pull that all in separately."Kareem Dossa, demo call, 5/28/2026 (32:14)
"I want our data protected especially as we, we divulge… that's the magic here is connecting signals or triggers to a relationship."Kareem Dossa, demo call, 5/28/2026 (28:42)
3 conversations · 2 transcripts (Fathom 5/22 + Hub DB 5/28 id 2d681e10-…) · 834 sentences · 17 chunks · 13 action items · all queryable.
Connecting signals or triggers to a relationship — that's the magic here.
This is the architecture under Stages 1 and 2 — the core paid engagement. The output is what Kareem asked for: a daily prioritized "qualified opportunity list" each broker walks into the day with — signal on the left, relationship hook on the right, recommended persona & first message attached.
Signals on their own are noise. The full engine weighs every live signal against standing documentation on every account — funding history, lease history, scientific platform, growth trajectory, decision-maker fit. That weighted score surfaces the right signal to the right rep at the right time.
Same math as the methodology Kareem already runs in his head. Faster, cheaper, at 18-broker scale.
Every artifact below lives in our Hub database with embeddings, ready to be RAG-retrieved by an agent on Colliers' behalf. None of this is for show — it's the corpus the system would draw from on day one.
Track A pipeline pre-5/22 — domain reports, frameworks, deep research from Gemini, Claude, Perplexity.
6 domain reports + 4 frameworks + 8 Gemini Deep Research topics + 40-company Bay Area life-sci universe.
Auto-extracted from the 5/28 transcript by speaker — Kareem's commitments, Brian's deliverables, Anis's coordination.
All retrievals are RAG-driven from Hub DB. See how the knowledge graph works → Plus: 2 transcripts ingested (834 sentences, 17 chunks), 11 brand tokens captured (this entire site is rendered in them), 8 Gemini Deep Research runs on PropTech + life-sci ecosystem partnerships.
Most CRMs go silent after close. Stage 7 is where margins live; Stage 1 is where the next deal starts. Same engine. Same Gold Sheets. Same relationships.
Kareem mentioned a Carnegie Mellon student team building a generic broker AI tool with three pillars. The pillars are right. The methodology is generic. We deliver the same three pillars but life-sci-tuned, Champions-aware, and built to Kareem's funnel — not a one-size-fits-all CRE tool.
CMU: exec moves, lease expirations, growth triggers.
BC: all that, plus IND filings, SEC 8-Ks, NIH grants, patents, and 6 Bay Area county lease-memo feeds. Life-sci-native day one.
CMU: ranks and prioritizes outreach.
BC: same, but stage-aware per Kareem's persona-by-stage logic + relationship-strength weighted + Miller-Heiman-aware via Gold Sheets (details).
CMU: alerts on stale relationships.
BC: same, plus the closed-loop circle — Gold Sheets persist beyond close, expansion signals route back to original rep with full context loaded.
A custom engagement, not the standard BioCreative SKU. Kareem's been clear he doesn't want to build it himself and doesn't want a generic platform — so we'd build it for him, with explicit ownership transfer at handoff. Four tracks. We're not laying out weeks or dates here on purpose — June 5 is where we discuss scope, collaboration model, your team's involvement, and what tracks you actually want first.
Tracks aren't strictly sequential. With the right collaboration model, Foundation and Engine can run in parallel. June 5 is when we land on the right shape. No prices on this page on purpose — depends on (a) which CRE data Colliers licenses, (b) network-health pillar scope, (c) how the work gets divided, (d) which optional extensions you bundle.
This is the substance of the conversation. Not a re-pitch. Honest unknowns — we don't pretend to have answers we don't have. Each question maps to something on the site we couldn't decide on without you in the room.
Does Colliers already license CoStar, Reonomy, or CompStak — and at what tier?
Drives whether the lease-data layer is "plug in your seat via API" or "we add ~$1.5K–4K/mo." Affects price more than any other variable. (CRE stack details)
Is there a Colliers global data team or AI initiative that already runs some of these feeds enterprise-wide?
If yes — we integrate, not duplicate. (Platform architecture)
What's the current state of Kareem's Champions DB and Trusted Advisors DB? Spreadsheet? Salesforce custom objects? In-flight build?
Drives Foundation-track scope. We need the schema in hand to design the matcher correctly.
How does the Active Pursuit process work today across the 18 brokers? Spreadsheet? Salesforce reports? Per-broker ad hoc?
We need the AS-IS workflow before designing the TO-BE.
What's an acceptable signal latency for Colliers? Same-day surfacing? 24-hour digest? Weekly batch?
Drives infrastructure choices and ongoing data-feed costs.
Scott — does this decision sit with you and Kareem, or is there a regional Colliers exec above you for AI/data spend?
If escalation is needed, we want to know now.
How does your team want to be involved in the build vs. have us deliver? Embedded engineer alongside Kareem's BI lead, or arms-length delivery with checkpoints?
Drives how the tracks sequence and how much your team learns vs. inherits.
Pilot with 3–4 brokers first, or 18-broker rollout from day one?
Pilot lowers risk; full rollout shows commitment. Kareem on 5/28 leaned pilot.
Confirm again — outbound runs on Colliers' systems, managed by BC? Including LinkedIn from broker accounts & email from Colliers domain?
Kareem said yes on 5/14 in writing. Want Scott aligned. IT/security implications.
Per-tenant DB? Per-tenant schema in shared DB? On Colliers infra or BC infra during the build?
Kareem flagged this unprompted on 5/28. Drives architecture before anything else. (platform isolation model)
Tier A consulting · Tier B dev support · Tier C managed-service-with-SLA — which is the right fit?
Drives the recurring-revenue model post-handoff.
Build the relationship-health monitor as part of Track 3 / Activation, or save for a later round?
If yes, that's the differentiator vs. CMU and the most defensible long-term moat.
Of the extensions across stages 3, 4, and 7 — sales support, legal, expansion + retention, plus the marketing stack — what's interesting? What's not?
Helps us size the right initial scope and what to roadmap as future phases.
