Saka One Portfolio
Saka One Enterprises 2026, Pilot v1

CRE Deal Scenario Platform.
From OM to Indicative Scenario in NBD

Commercial debt pre-screening tool for brokers and investors. AI extraction of OM, T-12, and rent roll into a credible indicative financing scenario. Operator in the loop on confidence scoring before the rules engine becomes autonomous.

7 steps
Intake to delivery workflow with SLAs
NBD
Target turnaround (next business day)
3 tiers
Confidence scoring (CONFIRMED, FLAGGED, MISSING)
2 lanes
Systems + Product / Market + Domain partnership
The Problem

Quote Lag Kills Deal Momentum

Getting a soft quote on a commercial real estate deal takes 1 day to 1 week. That lag has real costs. Brokers and investors wait days for indicative financing terms, long after the moment of maximum deal interest. Offering Memoranda (OMs) vary widely in structure, completeness, and how financials are presented. Pro forma versus actual is often unclear.

By the time a quote arrives, momentum is gone. Deals fall through. Buyers move on. Financing conversations start too late. The fix is not a faster human. It is a smarter intake workflow with immediate indicative output.

Why It Works

Six Structural Advantages from Day One

This product has the operating conditions most early-stage builds lack. The user, the source document, the pain, and the distribution path are all known.

01

Defined user

Brokers and investors. Known population, shared vocabulary, common workflow.

02

Known source document

The OM is a standard entry artifact across CRE. Consistent starting point for extraction.

03

Existing pain

No need to sell the problem. Users already feel the slow-quote friction every week.

04

Warm distribution

Initial access through trust, not cold outreach. Rare early advantage.

05

Embedded monetization

Every scenario that converts is a self-qualified lead with deal context attached.

06

Real feedback from day one

Testing on live deals means the feedback loop starts before the build is done.

7-Step Intake to Delivery Flow

Operator in the loop on confidence scoring. Target turnaround is next business day. Same-day delivery for simple text-based OMs received before noon.

1
Receive deal
User emails docs to intake address or drops in shared folder. Check completeness. Email user within 2 hours of receipt with status.
15 min
2
Run extraction
Upload PDF to extraction tool. Run structured extraction prompt. Review raw JSON output for obvious errors before scoring.
20 to 30 min
3
Score confidence
For each critical field, assign CONFIRMED, FLAGGED, or MISSING per the checklist. Do not skip this step.
20 min
4
Review NOI (cannot be skipped)
Confirm whether NOI is trailing actual or pro forma. The single non-delegable step. Escalate to partner if pro forma only.
10 min
5
Assemble output
Populate the deal summary template with confirmed fields. Mark flagged fields. Populate action list with missing items.
20 min
6
Deliver
Email output PDF to broker. CC operator. Subject line standardized.
5 min
7
Log feedback
Capture broker accuracy ratings in Airtable. Rate each field Correct, Off, or Wrong. Note deal-specific issues.
15 min

Confidence Scoring Framework

Three explicit tiers. Never present a flagged or computed figure without labeling. Never present a missing field as if it has a value.

CONFIRMED
Field explicitly stated in the document AND consistent with at least one other document or section. Safe to present as fact.
FLAGGED
Field stated in only one location with no cross-reference, OR computed/inferred from other fields. Must be labeled as flagged in output.
MISSING
Field not present in any provided document. Goes to the action list. Never left blank in output.

Escalation Rules. Two situations always require partner sign-off

Some failure modes cannot be handled by the operator alone. The system escalates explicitly:

  • NOI as pro forma only. If NOI is only available as a pro forma figure with no trailing actual in any provided document, escalate before delivery.
  • Cross-document discrepancy greater than 10%. If any critical field shows a discrepancy of more than 10% between sources with no explanation, escalate before delivery.
  • Scanned image OM. If the OM is a scanned image PDF (not text-based), do not run standard extraction. Flag, ask broker for text version, or run OCR with low-confidence labeling on all fields.

Two-Lane Partnership Model

Solo build with industry expert distribution. Lanes are explicit so neither side assumes contributions. Domain partner's ongoing contribution keeps lender logic current and product credible over time.

Lane 1
Systems + Product (Mike)
  • Workflow design and product architecture
  • Document parsing and intake logic
  • Rules engine and scenario generation
  • UI/UX and screen-by-screen flow
  • CRM, lead routing, feedback loops
  • Operational scaling infrastructure
Lane 2
Market + Domain (Partner)
  • Market truth and lender vocabulary
  • Credibility with pilot users
  • Testing on live deals
  • Ongoing refinement of quote logic
  • Distribution through relationships
  • CRE pattern recognition
What This Demonstrates

The Pattern Behind the Product

CRE Deal Scenario Platform applies the same operating discipline visible in every prior engagement. Manual MVP first. Confidence scoring as an explicit framework, not buried in product behavior. Operator in the loop on the steps that cannot afford to be wrong. Clear escalation rules. Two-lane partnership where each side owns what they own and nothing they do not.

The path from this v1 to the autonomous rules engine is straight. Every operator decision becomes training data. Every broker rating in Airtable becomes a model improvement signal. Every escalation refines the rule set. The manual workflow is the spec.

Back to Case Studies

Amazon Ads

From Fragmented Events to a Governed Commercial Operating System

Read the case