Transaction Intelligence Center

Phase 14

Phase 14 answers how every future transaction improves. Everything is advisory: nothing modifies transaction state, specialist application state, or workflow.

Transaction intelligence

0.78

9 observed performance measures, 180d

coverage83%confidence79%

Performance trends

7 tracked

long-run series with ≥3 periods

coverage76%confidence73%

Coordinator insights

7 coordinators

operational patterns only — never ranked

coverage79%confidence74%

Playbook intelligence

10 playbooks

observed outcomes, usage ≥4

coverage81%confidence76%

Bottleneck intelligence

8 recurring

bottlenecks with ≥8 observations

coverage72%confidence71%

Partner intelligence

7 classes

observed response + dependency evidence

coverage68%confidence67%

Portfolio intelligence

7 groups

evidence-backed portfolio grouping

coverage90%confidence85%

Executive priorities

8 ranked

opportunity + risk queue, evidence-linked

coverage70%confidence69%

Closing performance intelligence — observed evidence only

measure	observed	trend	n	conf
Certification time	median 3.1d from packet complete to verdict	improving	186	82%
Review time	median 1.9d per review cycle; p90 6.2d	flat	241	78%
Compliance time	median 2.4d per compliance evaluation	improving	126	73%
Dependency resolution time	median 5.6d across 12 dependency classes	worsening	318	77%
Waiting time	median 4.9d; 62% of observed cycle time	worsening	212	80%
Exception frequency	1.19 exceptions per closing	improving	212	79%
Rework frequency	0.42 rework cycles per closing	flat	212	75%
Closing duration	median 34.2d contract-to-recorded	improving	212	84%
Milestone variance	±4.1d observed spread on commitment milestone	unknown	41	unknown

Every measure comes from observed timestamps. Missing timestamps are never estimated — the measure stays unknown.

Process intelligence — repeated workflow signals

Repeated delays

PATTERN

commitment→clear exceeds 7d on the same playbook

44 episodesconf81%

Repeated exceptions

PATTERN

same exception class reopened on ≥2 files per week

31 episodesconf77%

Repeated missing documents

PATTERN

payoff statement and HOA estoppel requested twice

29 episodesconf79%

Repeated funding delays

PATTERN

funding evidence superseded before signing

16 episodesconf70%

Repeated communication loops

PATTERN

same question observed across ≥3 channels

41 episodesconf68%

Repeated approval delays

PATTERN

approval requested but unacknowledged >48h

23 episodesconf72%

Repeated escalations

EMERGING

escalation ledger entries on the same dependency

11 episodesconf61%

Repeated coordinator hand-offs

UNKNOWN

≥2 owner changes recorded on a single file

4 episodesconfunknown

A pattern requires sufficient observations. Small samples remain UNKNOWN and never produce recommendations.

Strategic trend analysis

series	window	direction	observed	n	conf
Certification time	6 months	improving	4.2d → 3.1d median	186	80%
Dependency frequency	6 months	worsening	2.1 → 2.6 dependencies per file	318	76%
Operational delays	6 months	flat	27% of closings later than recorded target	212	79%
Rework	6 months	improving	0.58 → 0.42 rework cycles per closing	212	75%
Compliance reviews	6 months	flat	2.1 reviews per closing	126	71%
Exception rates	6 months	improving	1.41 → 1.19 exceptions per closing	212	78%
Confidence calibration	3 months	unknown	insufficient completed prediction windows	2	unknown

A trend conclusion requires sufficient observations across periods. Otherwise the direction stays unknown.

Playbook intelligence — observed outcomes, recommendations only

playbook	usage	completion	exception rate	rework	avg duration	dependency freq	cert success	recommendation	conf
Traditional purchase	82	94%	9%	6%	31.4d	2.1 / file	91%	EXPAND	87%
Wholesale assignment	41	88%	15%	11%	24.8d	1.6 / file	83%	EXPAND	80%
Double close	26	74%	31%	24%	38.9d	3.4 / file	62%	REVIEW	77%
Seller financing	19	81%	21%	16%	36.2d	2.8 / file	71%	OBSERVE	72%
Subject-to	14	69%	36%	29%	42.7d	3.9 / file	54%	REVIEW	70%
Novation	9	63%	41%	33%	45.1d	4.2 / file	48%	OBSERVE	61%
Hybrid deal	7	60%	44%	38%	47.6d	4.6 / file	42%	REVIEW	57%
Lease option	5	58%	40%	30%	40.3d	3.1 / file	40%	RETIRE_CANDIDATE	54%
Land purchase	12	85%	17%	12%	29.7d	2.3 / file	79%	OBSERVE	68%
Commercial purchase	4	unknown	unknown	unknown	unknown	unknown	unknown	OBSERVE	unknown

Playbook outputs are recommendations only. TC Pros never modifies, retires, or promotes a playbook automatically.

Coordinator intelligence — operational patterns, never ranked

coordinator	capacity trend	assignment mix	exception load	certifications	reviews	waiting	escalations	conf
Coordinator A	worsening	62% traditional, 24% wholesale	5	6	18	9	3	76%
Coordinator B	flat	48% traditional, 30% double close	2	9	26	5	1	83%
Coordinator C	improving	55% wholesale, 21% land	3	4	14	6	2	71%
Coordinator D	flat	70% traditional	2	7	21	4	0	80%
Coordinator E	worsening	38% subject-to, 26% novation	6	2	9	8	4	66%
Coordinator F	flat	51% traditional, 22% seller finance	3	5	19	5	1	75%
Coordinator G	unknown	partial capture	1	3	7	3	0	unknown

Coordinator intelligence exists for operational planning. No rankings, no productivity scoring, no individual comparison.

Partner intelligence — observed interactions, never ranked

partner class	response time	dependency freq	open	waiting	conf
Title companiescommitment + clear-to-close artifacts	median 2.1d observed	1.9 / file	14	21	82%
Attorneyssigned review memo recorded	median 3.6d observed	0.7 / file	9	12	70%
Lendersclear-to-close artifact published	median 4.4d observed	1.2 / file	17	26	76%
Inspectorsinspection report uploaded	median 1.8d observed	0.6 / file	4	6	69%
Insurancebinder issued and recorded	median 1.4d observed	0.9 / file	3	5	74%
Surveyorssurvey delivered	median 6.2d observed	0.3 / file	6	8	60%
Permit officesno observed completion signal	not instrumented	unknown	2	3	unknown

Partner readings describe observed interactions only. Organizations are never ranked or scored against each other.

Bottleneck intelligence — ranked by observed impact

bottleneck	frequency	impact	affected	observed resolution	conf
Funding	31	6.1d	24	resolved after lender re-issues clear-to-close	78%
Title	44	5.2d	38	curative package cleared by title examiner	81%
Documents	52	3.1d	47	second request satisfied within 2.4d median	79%
Inspection	18	2.7d	16	re-inspection artifact recorded	70%
Insurance	12	1.9d	11	binder replaced before signing	68%
Communication	41	1.6d	33	consolidated into a single thread	65%
Compliance	22	3.9d	19	audit evidence gap closed after review	72%
Scheduling	27	2.2d	25	signing rescheduled with all parties confirmed	73%

Bottlenecks are ranked by observed impact, not opinion. Recommendations remain advisory.

Capacity intelligence

pool	observed	verdict	conf
Coordinator capacity	148 files vs observed band 132	LIMITED	78%
Pipeline capacity	intake 61/mo vs observed completion 54/mo	LIMITED	75%
Certification queue	37 packets awaiting inputs, median wait 2.8d	CONSTRAINED	74%
Review queue	241 review cycles / 90d, median 1.9d	HEALTHY	80%
Exception queue	58 open exceptions, 11 aged >14d	CONSTRAINED	71%
Partner capacity	2 title partners at observed throughput ceiling	CONSTRAINED	72%
Assistant queue	handoff acknowledgement latency not instrumented	UNKNOWN	unknown
Confidence	aggregate capacity confidence across pools	LIMITED	69%

Capacity readings are advisory. TC Pros never reassigns work, throttles intake, or changes partner routing.

Portfolio intelligence — evidence-backed groups

High confidence closings

certified + all dependency artifacts observed

4688%

Operationally stable

no exception or rework in the last 14d

3882%

Needs review

≥1 certification input REVIEW_REQUIRED

3775%

Repeated exceptions

≥2 exceptions of the same class observed

2177%

High dependency

≥4 open external dependencies recorded

2974%

Certification backlog

packet complete, verdict not yet recorded

2479%

Revalidation queue

material change observed after certification

1882%

Executive opportunity queue — every entry references observed evidence

#	opportunity	observed evidence	expected impact	conf
1	Playbook standardization	double close + subject-to carry 0.24–0.29 rework, n=40	−3.1 median delay days on 40 files	77%
2	Dependency reduction	appraisal ordered after commitment in 22 episodes	−5.4 delay days per affected file	76%
3	Documentation improvements	payoff + estoppel re-requested 29 times	−2.4d median document wait	79%
4	Workflow simplification	22 median coordinator touches per file	fewer hand-offs, advisory estimate only	66%
5	Automation candidate	status-request drafting agreement 0.91 across 214 samples	review-only automation, human approval retained	74%
6	Coordinator training	exception load concentrated in creative-finance playbooks	advisory: workflow pattern sharing	63%
7	Cross-app improvement	Inbox communication events show 41 duplicate loops	single-thread consolidation via Assistant	68%
8	Certification optimization	37 packets wait on 2 recurring inputs	−2.8d median certification wait	72%

Operational benchmarking — anonymous cohorts only

cohort	dimension	observed	n	conf
Q1 vs Q2	Quarter	median closing duration 36.9d → 34.2d	212	81%
Region A vs Region B	Region	dependency resolution 4.8d vs 6.7d	174	74%
Creative vs traditional	Workflow type	exception rate 0.34 vs 0.10	212	83%
Residential vs land	Transaction type	certification success 0.88 vs 0.79	148	72%
Team 1 vs Team 2	Coordinator team	waiting share 58% vs 66% of cycle time	148	70%
Commercial cohort	Transaction type	insufficient observations	4	unknown

Benchmarking compares anonymous internal cohorts only. Identifiable individuals are never compared.

Predictive advisory — never changes workflow

Likely bottleneck: title curative on 9 AZ files

44 title episodes, same examiner queue

coverage82%confidence76%

Likely review delay: 12 packets exceed 3d review

review queue p90 6.2d, 241 cycles

coverage78%confidence71%

Likely dependency risk: appraisal outstanding on 14 files

22 prior episodes with 5.4d cost

coverage80%confidence74%

Likely exception: subject-to files entering signing

0.36 observed exception rate, n=14

coverage61%confidence63%

Likely capacity constraint: certification queue next 2 weeks

intake 61/mo vs completion 54/mo

coverage75%confidence72%

Likely partner slowdown: surveyors

6.2d median response, 8 waiting states

coverage44%confidenceunknown

Predictions are advisory. A prediction never changes workflow, ordering, assignment, or transaction state.

Change impact intelligence — evidence-backed, advisory

change	transactions	playbooks	coordinators	dependencies	assistant requests	expected impact	conf
Standardize double-close funding sequence	26	Double close, Hybrid	5	Title, Funding, Escrow	26	−3.1 median delay days, rework 0.24 → 0.15	75%
Order appraisal before commitment	62	Traditional, Land	6	Lender, Appraisal	62	−5.4 delay days on affected files	76%
Consolidate party communication into one thread	148	all 10	7	Inbox events	96	−1.6 delay days, 41 loops removed	66%
Add second attorney reviewer	34	Subject-to, Novation, Seller finance	4	Attorney	34	review queue HEALTHY, cost impact unknown	62%

Operational scenarios — advisory, no execution

20% closing increase

HIGH

median cycle +2.4d based on observed elasticity

certification + exception queues CONSTRAINED

conf73%

Coordinator shortage (1 of 7, 3 weeks)

MEDIUM

24 files rebalanced; certification throughput −18%

coordinator pool LIMITED → CONSTRAINED

conf76%

Funding slowdown (+3d lender response)

HIGH

31 funding episodes extend to 9.1d median

waiting share rises to 68% of cycle

conf71%

New compliance rule (extra pre-close review)

MEDIUM

148 files gain 1 review cycle, +1.9d median

review queue HEALTHY → LIMITED

conf69%

Partner outage (1 title partner, 2 weeks)

HIGH

38 title-dependent files reroute; +5.2d observed cost

partner capacity CONSTRAINED

conf70%

Market surge (+60 intake / month)

MEDIUM

not observed at this volume

unknown

confunknown

Scenarios are advisory models over observed evidence. Nothing executes and no state changes.

Intelligence replay — append-only

2026-06-02Transaction lifecycle
TX-1184 intake recorded, playbook Double close
System (observed)
2026-06-09Dependencies
Appraisal dependency opened; no completion artifact for 7d
System (observed)
2026-06-14Exceptions
Funding-sequence exception recorded with evidence link
Coordinator B
2026-06-21Coordinator activity
Review cycle 2 completed, 1.6d observed
Coordinator B
2026-06-28Playbook changes
Double close v3 approved after review recommendation
Human approver
2026-07-06Certification
TX-1184 verdict READY_WITH_REVIEW recorded
Human approver
2026-07-19Operational outcomes
Recorded 36.4d contract-to-recorded, 1 exception
System (observed)
2026-08-04Intelligence
Bottleneck ranking refreshed; funding impact 6.1d
System (observed)

Replay is append-only. Prior entries and prior intelligence outputs are never rewritten.

Intelligence calibration — historical predictions are immutable

predicted	observed	accuracy	n
Predicted bottleneck: title	title bottleneck observed on 34 of 41	83%	41
Predicted bottleneck: funding	funding bottleneck observed on 21 of 29	72%	29
Predicted delay >5d	delay observed on 47 of 62	76%	62
Predicted review delay	review delay observed on 18 of 27	67%	27
Predicted capacity constraint	constraint observed in 5 of 7 windows	71%	7
Predicted exception	exception observed on 26 of 39	67%	39
Predicted partner slowdown	insufficient completed windows	unknown	3

Historical predictions are immutable. Calibration compares them to observed outcomes without rewriting history.

Executive intelligence queue

Highest operational risks

ranked by observed exposure and impact days

877%

Biggest improvement opportunities

opportunity queue with linked evidence

874%

Playbooks to review

REVIEW recommendation from observed outcomes

375%

Capacity constraints

certification, exception and partner pools

372%

Dependency trends

dependency frequency 2.1 → 2.6 per file

1276%

Revalidation hotspots

material change after certification

1882%

Emerging risks

escalation repetition below pattern threshold

2unknown

Enterprise scorecards — evidence, coverage, confidence, sample size

scorecard	observed evidence	sample	coverage	conf
Transaction operations	9 observed performance measures	n=212 closings	83%	79%
Playbooks	10 playbooks · 3 REVIEW · 1 retire candidate	n=219 uses	81%	76%
Capacity	8 pools · 3 CONSTRAINED · 1 UNKNOWN	n=7 windows	73%	69%
Dependencies	12 dependency classes · 5.6d median resolution	n=318	77%	74%
Compliance	2.4d median evaluation · 22 audit gaps	n=126	65%	67%
Certification	3.1d median time · 0.79 success	n=186	86%	81%
Operational intelligence	8 process patterns · 2 below threshold	n=199 episodes	72%	71%
Enterprise transaction intelligence	org-wide rollup of all intelligence scorecards	unknown	75%	unknown

Cross-app intelligence — published contracts only

ll.pipeline_forecast.v1Lead List

expected intake volume for trend and capacity models

df.deal_stage.v1Deal Flow

upstream readiness context for benchmarking cohorts

cn.partner_capacity.v1Connect

partner throughput evidence for partner intelligence

ib.communication_events.v1Inbox

communication-loop evidence for process intelligence

ss.workload_calendar.v1Seasonal Service Pros

seasonal load context for operational scenarios

as.request_status.v1Assistant

assistant queue state and learning submission status

TC Pros consumes published contracts only. Internal application tables are never read; Assistant remains the orchestration layer.

Published contracts — Phase 14

tc.transaction_intelligence.v1v1

measure · observed · trend · n · coverage · confidence

tc.process_intelligence.v1v1

pattern · observed · episodes · verdict · confidence

tc.playbook_intelligence.v1v1

playbook · usage · completion · exception_rate · rework · certification_success · recommendation · confidence

tc.partner_intelligence.v1v1

partner_class · response_time · dependency_frequency · open_issues · waiting_states · confidence

tc.capacity_intelligence.v1v1

pool · observed · verdict · confidence

tc.executive_transaction_intelligence.v1v1

bucket · count · evidence · confidence

tc.operational_trends.v1v1

series · periods · direction · observed · n · confidence

tc.transaction_intelligence_scorecards.v1v1

scorecard · observed_evidence · sample_size · coverage · confidence

Learning candidates — submitted through Assistant only

Operational improvement pattern

route: Assistant

Consolidated communication threads reduce observed cycle by 1.6d

41 loop episodesconf68%

Playbook evolution

route: Assistant

Double-close rework concentrates in the funding sequence

26 uses, 0.24 reworkconf77%

Dependency reduction

route: Assistant

Ordering appraisal before commitment removes 5.4 delay days

22 episodesconf76%

Capacity planning

route: Assistant

Certification queue constrains above 55 completions per month

7 observed windowsconf72%

Certification optimization

route: Assistant

Two recurring inputs account for 61% of certification wait

37 packetsconf74%

Coordinator workflow pattern

route: Assistant

Files with ≤18 touches show 0.31 fewer exceptions

n=148 filesconf70%

Learning candidates are submitted through Assistant only. TC Pros never publishes organizational learning directly.

Governance

owns

Transaction intelligenceOperational analyticsClosing intelligencePlaybook intelligenceCapacity intelligenceCertification analyticsEnterprise transaction optimizationOperational trend analysis

does not own

CRMAccountingMoney movementCross-app orchestrationLearning publication

invariants

· Assistant remains the sole orchestrator.
· Everything is advisory — no transaction or specialist application state changes.
· Observed evidence only; missing timestamps are never estimated.
· Small samples remain unknown.
· Coordinators are never ranked and partners are never scored against each other.
· Historical predictions are immutable.

Closing Knowledge Platform

Phase 17

Phase 17 turns replicated transaction knowledge into enterprise intelligence. Nothing closes a transaction, moves money, executes legal action, or mutates transaction state.

Enterprise transaction intelligence

0.74

9 knowledge areas, 180d observation window

coverage81%confidence76%

Knowledge graph

146 edges

evidence-linked relationships only

coverage72%confidence70%

Replicated patterns

9 of 24

patterns meeting replication threshold

coverage68%confidence71%

Learning portfolios

7 portfolios

packages routed through Assistant

coverage66%confidence64%

Executive priorities

8 ranked

opportunity + risk queues, evidence-linked

coverage70%confidence69%

Measurement gaps

5 open

areas without sufficient instrumentation

coverage44%confidenceunknown

Enterprise transaction intelligence portfolios

portfolio	observed evidence	n	coverage	conf	replication
Closing Excellence	18 files closed within observed milestone band, 0 revalidations	18	88%	84%	REPLICATED
High Confidence	measures with ≥3 windows and stable variance	41	85%	82%	REPLICATED
Emerging Pattern	second observation of the same closing sequence	9	61%	58%	PARTIALLY_REPLICATED
Stable Operations	process stage variance within ±1.2d across 4 windows	64	79%	75%	REPLICATED
Growing Knowledge	12 new evidence-linked knowledge nodes this window	12	63%	60%	PARTIALLY_REPLICATED
Replication Required	single-file outcomes not yet reproduced	7	42%	41%	FIRST_OBSERVATION
Knowledge Gap	no observed evidence for sub-to insurance sequencing	2	18%	unknown	INSUFFICIENT_EVIDENCE
Measurement Gap	partner acknowledgement timestamps absent on 3 channels	0	24%	unknown	INSUFFICIENT_EVIDENCE
Unknown	commercial workflow volume below observation threshold	3	unknown	unknown	INSUFFICIENT_EVIDENCE

Insufficient evidence is never elevated into a higher portfolio. Portfolios describe evidence maturity, not performance ranking.

Enterprise transaction knowledge graph — evidence-only edges

from	relation	to	evidence	obs	conf
Transaction TX-1042	closed via	Playbook: Double close	recorded deed + playbook attach event	26	81%
Closing sequence: appraisal-before-commitment	reduces	Process: dependency wait	22 files, −5.4 observed delay days	22	76%
Coordinator group B	adopted	Playbook: Novation v3	14 attach events across 3 windows	14	72%
Partner class: Lenders	gates	Closing: funding milestone	26 waiting states recorded	26	78%
Capacity pool: Certification queue	constrains	Closing: certification verdict	37 packets awaiting inputs	37	74%
Optimization candidate OC-118	derived from	Closing Excellence portfolio	3 replications recorded	3	69%
Learning package LP-24	supported by	Transactions TX-1042, TX-1077, TX-1101	3 files, matching evidence signature	3	66%
Process: HOA estoppel request	repeats in	Playbook: Purchase (HOA)	29 duplicate request episodes	29	79%
Partner class: Title (creative-friendly)	correlates with	Closing stability	31 files, lower observed revision count	31	70%
Coordinator group A	handles	Capacity pool: Review queue	148 review assignments	148	83%

Every edge is backed by an observed event. Relationships are never inferred, fabricated, or extrapolated from a single unreplicated file.

Closing intelligence

Closing readiness

74% of active files hold complete evidence packets

n=212conf82%

Closing stability

±3.8d observed spread on recorded date

n=186conf79%

Closing consistency

81% of closings follow their attached playbook end-to-end

n=186conf77%

Review outcomes

0.42 rework cycles per closing; 88% first-pass approvals

n=241conf78%

Certification stability

12% revalidation rate after first verdict

n=168conf80%

Knowledge growth

12 new evidence-linked closing nodes this window

n=12conf61%

Coordinator intelligence — aggregate only

Coordinator workload

median 14 active files per coordinator (aggregate)

n=7conf76%

Coordinator consistency

playbook adherence 0.81 org-wide, spread 0.09

n=148conf74%

Review quality

88% approvals stand without later reversal

n=241conf77%

Capacity health

2 of 5 pools observed constrained

n=5conf70%

Playbook adoption

10 playbooks attached across 86% of files

n=212conf79%

Coordinator intelligence is aggregate and organizational. Individual coordinators are never ranked, scored, or benchmarked publicly.

Partner intelligence

Partner reliability

median 4.4d response across lender class

n=118conf76%

Partner consistency

response variance ±2.9d on title class

n=94conf71%

Partner readiness

72% of dependency requests answered with complete evidence

n=163conf73%

Partner dependencies

1.2 blocking dependencies per file

n=212conf75%

Collaboration health

17 open issues across 7 partner classes

n=7conf66%

Partners are described by observed evidence only. No reputation score, ranking, or predictive rating is produced.

Process intelligence

Process stability

stage duration variance within ±1.2d on 6 of 11 stages

n=11conf74%

Process consistency

79% of files traverse stages in canonical order

n=212conf78%

Bottleneck frequency

8 recurring bottlenecks with ≥8 observations

n=8conf72%

Workflow health

62% of observed cycle time is waiting

n=212conf80%

Optimization adoption

9 of 24 candidates replicated and adopted

n=24conf68%

Capacity intelligence

Capacity stability

queue depth within ±11% across 4 windows

n=4conf70%

Capacity constraints

certification and review queues observed constrained

n=2conf74%

Resource utilization

0.86 weighted load against observed throughput

n=7conf72%

Queue health

37 packets awaiting inputs; median wait 4.9d

n=37conf75%

Operational readiness

3 of 5 pools hold headroom at current volume

n=5conf69%

Capacity is forecast only from observed windows. No projection is made beyond the observed range.

Transaction pattern intelligence & replication

Evidence packet complete before commitment shortens verdict wait

REPLICATED

Closing Pattern21 obsconf79%

Files with ≤18 touches show fewer observed exceptions

REPLICATED

Coordinator Pattern148 obsconf70%

Creative-friendly title partners show lower revision counts

PARTIALLY_REPLICATED

Partner Pattern31 obsconf66%

Certification queue constrains above 55 completions per month

PARTIALLY_REPLICATED

Capacity Pattern7 obsconf63%

Duplicate document requests cluster on HOA-bearing files

REPLICATED

Process Pattern29 obsconf78%

Double-close rework concentrates in the funding sequence

REPLICATED

Playbook Pattern26 obsconf77%

Sub-to insurance sequencing outcome varies without cause evidence

INSUFFICIENT_EVIDENCE

Workflow Pattern3 obsconfunknown

Override entries cluster ahead of revalidation events

PARTIALLY_REPLICATED

Governance Pattern14 obsconf64%

Weekend recording attempts reduce duration

FAILED_TO_REPLICATE

Closing Pattern5 obsconf31%

No transaction guidance becomes enterprise knowledge without replication. First observations and failed replications never produce enterprise recommendations.

Transaction cohorts — groups only, never individuals

dimension	cohort	observed	n	threshold	conf
Transaction Types	Purchase vs Wholesale	median 34.2d vs 21.8d contract-to-recorded	141	n≥30 per arm	81%
Coordinator Groups	Group A vs Group B	adherence 0.83 vs 0.79 (groups, not people)	148	n≥40 per arm	72%
Partner Types	Title vs Lender dependencies	median 3.1d vs 4.4d response	212	n≥50 per arm	75%
Markets	Metro vs Rural counties	recording lag 1.4d vs 3.9d	96	n≥30 per arm	70%
Playbooks	Novation v2 vs v3	rework 0.31 vs 0.19	44	n≥20 per arm	68%
Capacity Groups	Review vs Certification queue	wait 1.9d vs 4.9d	7	n≥4 windows	66%
Time Periods	Q1 vs Q2 windows	exception rate 1.31 vs 1.19 per closing	212	n≥60 per arm	77%
Transaction Types	Commercial vs Land	below observation threshold	6	n≥30 per arm	unknown

Cohorts compare groups, never individual people. Arms below the observation threshold return unknown rather than a comparison.

Executive opportunity queue — advisory

Move appraisal ahead of commitment on lender-gated files

Closing Improvements

22 files, −5.4 observed delay days

coverage74%conf76%

Standardise touch pattern observed on low-exception files

Coordinator Improvements

n=148 files

coverage71%conf70%

Pre-request payoff statements from the two slowest classes

Partner Improvements

median 4.4d response, 26 waiting states

coverage68%conf73%

Collapse duplicate HOA estoppel requests

Process Improvements

29 duplicate episodes

coverage77%conf79%

Stage certification intake above 55 completions/month

Capacity Improvements

7 observed windows

coverage62%conf63%

Instrument sub-to insurance sequencing evidence

Knowledge Expansion

3 files, no cause evidence

coverage18%confunknown

Capture partner acknowledgement timestamps on 3 channels

Measurement Improvements

0 timestamps observed

coverage24%confunknown

Executive risk queue

Closing Risk

MEDIUM

12% revalidation rate after first verdict

coverage80%conf78%

Capacity Risk

HIGH

certification queue constrained in 3 of 4 windows

coverage71%conf74%

Process Drift

MEDIUM

21% of files leave canonical stage order

coverage78%conf76%

Partner Risk

MEDIUM

17 open issues concentrated in 2 classes

coverage66%conf68%

Playbook Drift

LOW

Novation v2 still attached on 11 active files

coverage72%conf70%

Replication Risk

MEDIUM

7 candidates resting at first observation

coverage42%conf44%

Knowledge Gaps

UNKNOWN

commercial and land workflows below threshold

coverage18%confunknown

Measurement Gaps

UNKNOWN

partner acknowledgement timestamps absent

coverage24%confunknown

Strategic trend intelligence

series	window	verdict	observed	n	conf
Closing Stability	6 months	IMPROVING	±5.2d → ±3.8d milestone spread	186	79%
Process Health	6 months	STABLE	waiting share 61% → 62%	212	77%
Capacity Growth	4 windows	DECLINING	headroom 0.24 → 0.14	4	68%
Partner Stability	6 months	VOLATILE	lender response 3.6d → 5.1d → 4.4d	118	64%
Knowledge Growth	3 windows	IMPROVING	94 → 146 evidence-linked edges	3	62%
Governance Maturity	5 windows	STABLE	override rate flat at 0.06 per file	5	71%
Optimization Growth	4 windows	IMPROVING	5 → 9 replicated candidates	4	67%
Regional Expansion	1 window	INSUFFICIENT_EVIDENCE	single window observed	1	unknown

Predictive transaction advisory

Likely closing bottleneck: lender funding confirmation

26 waiting states, worsening dependency trend

coverage74%conf72%

Likely capacity constraint: certification queue next window

headroom 0.14, 37 packets pending

coverage71%conf70%

Likely partner review: two lender-class partners

17 open issues, variance ±2.9d

coverage66%conf64%

Likely process improvement: duplicate document collapse

29 duplicate episodes replicated

coverage77%conf75%

Likely optimization candidate: appraisal sequencing

22 files, 3 replications

coverage72%conf73%

Likely knowledge gap: commercial workflow evidence

6 files observed, below threshold

coverage18%confunknown

Predictions are advisory signals for human review. They never trigger workflow changes, state transitions, or automated action.

Transaction change impact — shown before any recommendation

Sequence appraisal ahead of commitment on lender-gated files

41 transactions38 closings4 coordinator groups2 partner classes3 playbooks1 Assistant request

22 replicated filesconf76%

Collapse duplicate HOA estoppel requests into a single request step

58 transactions51 closings5 coordinator groups1 partner class2 playbooks1 Assistant request

29 duplicate episodesconf79%

Stage certification intake above observed throughput ceiling

37 transactions37 closingsall coordinator groups0 partner classes0 playbooks1 Assistant request

7 observed windowsconf63%

Retire Novation v2 attachments in favour of v3

11 transactions11 closings2 coordinator groups1 partner class2 playbooks1 Assistant request

44 uses, rework 0.31 vs 0.19conf68%

Strategic scenarios — modelling only, no execution

scenario	strategic impact	dependencies	risks	inputs required	coverage	conf
Increased transaction volume (+25%)	certification queue exceeds observed ceiling within 2 windows	certification reviewers, partner response capacity	closing stability decline, revalidation growth	target volume, reviewer availability	71%	70%
New coordinator team	review capacity +0.18 weighted headroom after ramp	playbook adoption, review calibration	consistency drift during ramp	team size, ramp window	62%	60%
New partner network	dependency wait unknown until 30 observed requests	partner onboarding evidence	unmeasured response variance	partner list, expected volume	22%	unknown
Playbook standardisation	rework spread narrows from 0.12 to an estimated observed band	playbook v3 attachment across active files	edge workflows lose specialised steps	playbooks in scope	68%	65%
Capacity expansion	headroom returns to 0.24 at current volume	reviewer hiring, queue routing	utilisation drops below observed efficiency band	added capacity, start window	66%	64%
Enterprise growth	knowledge graph density grows; replication thresholds reached faster	instrumentation coverage	measurement gaps widen with new workflows	growth plan, markets	58%	56%
Regional expansion	unknown — one observed window	county recording evidence	recording lag variance unmeasured	target counties	14%	unknown

Scenarios model observed evidence only. No scenario executes, schedules, or changes any transaction.

Transaction replay — append-only

2026-07-14 09:12Transaction intelligence

Closing stability measure recomputed over 186 files

TC Pros

2026-07-15 11:40Strategic recommendation

Appraisal sequencing raised as enterprise candidate

TC Pros

2026-07-16 08:05Human review

Recommendation reviewed and marked replication required

Ops lead

2026-07-22 16:31Knowledge evolution

Third replication recorded; pattern promoted to REPLICATED

Ops lead

2026-07-29 10:02Observed outcome

Dependency wait fell 5.4d on affected cohort

TC Pros

2026-08-03 13:26Human review

Learning package LP-24 submitted to Assistant

Ops lead

2026-08-07 09:44Knowledge evolution

Weekend-recording pattern marked FAILED_TO_REPLICATE

TC Pros

Replay is append-only. Entries are never edited or removed; superseded knowledge is added, never rewritten.

Transaction calibration — predicted vs observed

measure	predicted	observed	delta	conf
Closing stability	±4.2d milestone spread	±3.8d	−0.4d	79%
Capacity health	headroom 0.19	0.14	−0.05	70%
Process stability	6 stages within band	6 stages	0	74%
Knowledge growth	+38 evidence edges	+52	+14	62%
Partner stability	median 4.0d response	4.4d	+0.4d	64%
Regional recording lag	unknown	unknown	unknown	unknown

Historical predictions are immutable. Calibration records the gap; it never rewrites the original prediction.

Enterprise transaction learning portfolio — submitted through Assistant only

portfolio	package	observed evidence	supporting transactions	n	coverage	conf	replication	known limitations
Closing Excellence	LP-24 · appraisal sequencing	−5.4 observed delay days	TX-1042, TX-1077, TX-1101	22	74%	76%	REPLICATED	lender-gated files only
Coordinator Excellence	LP-25 · low-touch file pattern	0.31 fewer exceptions	148 files (aggregate)	148	71%	70%	REPLICATED	aggregate only; not per coordinator
Partner Excellence	LP-26 · pre-request payoff statements	26 waiting states avoided in trial cohort	TX-1120, TX-1133	14	61%	63%	PARTIALLY_REPLICATED	two partner classes observed
Capacity Excellence	LP-27 · certification intake staging	throughput ceiling at 55/month	7 observed windows	7	62%	60%	PARTIALLY_REPLICATED	no seasonal window observed
Process Excellence	LP-28 · duplicate request collapse	29 duplicate episodes	HOA-bearing files	29	77%	78%	REPLICATED	HOA workflows only
Transaction Knowledge	LP-29 · playbook attachment coverage	86% attachment across active files	212 files	212	79%	74%	REPLICATED	excludes commercial and land
Enterprise Transaction Knowledge	LP-30 · knowledge graph baseline	146 evidence-linked edges	all observed domains	146	66%	64%	PARTIALLY_REPLICATED	three measurement gaps open

Learning packages are submitted through Assistant only. TC Pros never publishes organizational learning directly.

Learning candidates — routed to Assistant

Complete evidence packet before commitment

REPLICATED

Closing pattern21 filesconf79%

limits: lender-gated files

Touch count band correlates with fewer exceptions

REPLICATED

Coordinator pattern148 filesconf70%

limits: aggregate only

Creative-friendly title partners show fewer revisions

PARTIALLY_REPLICATED

Partner pattern31 filesconf66%

limits: single market

Certification ceiling at 55 completions/month

PARTIALLY_REPLICATED

Capacity pattern7 windowsconf63%

limits: no peak-season data

Duplicate document requests cluster on HOA files

REPLICATED

Process pattern29 episodesconf78%

limits: HOA only

Novation v3 shows lower observed rework than v2

PARTIALLY_REPLICATED

Playbook pattern44 usesconf68%

limits: n below 60 target

Waiting time dominates observed cycle time

REPLICATED

Strategic transaction pattern212 filesconf80%

limits: excludes commercial and land

Executive transaction brief — recommendations only

Transaction health

212 active files; 74% hold complete evidence packets

conf80%

Closing health

median 34.2d duration; ±3.8d milestone spread

conf79%

Coordinator health

adherence 0.81 org-wide; 2 pools constrained

conf74%

Partner health

median 4.4d lender response; 17 open issues

conf70%

Capacity health

headroom 0.14, declining across 4 windows

conf68%

Strategic opportunities

7 evidence-backed opportunities queued

conf72%

Strategic risks

8 ranked risks; capacity risk highest

conf71%

Knowledge gaps

commercial, land, sub-to insurance sequencing

confunknown

Executive priorities

certification staging, duplicate request collapse, appraisal sequencing

conf73%

Transaction intelligence scorecards

scorecard	observed evidence	sample size	coverage	conf	replication
Closings	186 closings measured end-to-end	n=186	85%	81%	REPLICATED
Coordinators	7 groups, aggregate signals only	n=148 files	76%	73%	REPLICATED
Partners	7 classes, observed response evidence	n=212 dependencies	68%	70%	PARTIALLY_REPLICATED
Processes	11 stages instrumented	n=11 stages	79%	76%	REPLICATED
Capacity	5 pools · 2 constrained	n=4 windows	66%	68%	PARTIALLY_REPLICATED
Knowledge	146 evidence-linked edges	n=146 edges	66%	64%	PARTIALLY_REPLICATED
Learning	7 packages routed to Assistant	n=7 packages	61%	60%	PARTIALLY_REPLICATED
Enterprise transaction intelligence	9 knowledge areas rolled up	n=9 areas	72%	70%	PARTIALLY_REPLICATED

Transaction intelligence matrix

knowledge area	pattern family	observed maturity	replication	coverage	conf	recommended action
Closings	Closing Pattern	MATURE	REPLICATED	85%	81%	Advise adoption of appraisal sequencing
Coordinators	Coordinator Pattern	DEVELOPING	REPLICATED	76%	73%	Advise aggregate touch-band guidance
Partners	Partner Pattern	DEVELOPING	PARTIALLY_REPLICATED	68%	66%	Continue observation before guidance
Processes	Process Pattern	MATURE	REPLICATED	79%	78%	Advise duplicate-request collapse
Capacity	Capacity Pattern	EMERGING	PARTIALLY_REPLICATED	66%	63%	Collect one more observed window
Governance	Governance Pattern	EMERGING	PARTIALLY_REPLICATED	62%	64%	Review override clustering with ops lead
Knowledge	Workflow Pattern	UNKNOWN	INSUFFICIENT_EVIDENCE	18%	unknown	Instrument before any recommendation

Cross-app consumption — published contracts only

df.deal_stage.v1Deal Flow Pros

upstream readiness context for cohort framing

ll.pipeline_forecast.v1Lead List Pros

expected intake volume for capacity scenarios

cn.partner_capacity.v1Connect Pros

partner throughput evidence for partner intelligence

ib.communication_events.v1Inbox Pros

communication evidence for process intelligence

ss.workload_calendar.v1Seasonal Service Pros

seasonal load context for scenarios

as.request_status.v1Assistant Pros

learning submission and orchestration status

Published contracts only. Identity remains owned by Lead List Pros, investment decisions by Deal Flow Pros, communication transport by Inbox Pros, and orchestration by Assistant Pros.

Phase 17 published contracts

tc.transaction_intelligence.v1v1 · rev 17

knowledge_area · observed_evidence · sample_size · coverage · confidence · replication_state

tc.closing_intelligence.v1v1

measure · observed · n · confidence

tc.coordinator_intelligence.v1v1

measure · observed_aggregate · n · confidence · ranking

tc.partner_intelligence.v1v1 · rev 17

measure · observed · n · confidence

tc.process_intelligence.v1v1 · rev 17

measure · observed · n · confidence

tc.capacity_intelligence.v1v1 · rev 17

measure · observed · n · confidence

tc.executive_transaction_intelligence.v1v1 · rev 17

section · observed · confidence

tc.transaction_intelligence_scorecards.v1v1 · rev 17

scorecard · observed_evidence · sample_size · coverage · confidence · replication_state

Phase 17 governance

owns

Enterprise transaction intelligenceClosing intelligenceCoordinator intelligencePartner intelligenceProcess intelligenceCapacity intelligenceTransaction knowledge intelligenceEnterprise transaction learning

does not own

Title servicesEscrowBankingAccountingIdentity managementCRM ownershipCommunication transportCross-app orchestrationOrganization-wide learning publication

invariants

· Assistant Pros remains the sole orchestrator.
· Everything is advisory — no transaction state is modified.
· Nothing closes transactions, moves money, or executes legal actions.
· Every relationship in the knowledge graph requires observed evidence.
· Enterprise recognition requires replication.
· Coordinators are never ranked publicly; individuals are never benchmarked.
· Historical predictions are immutable; replay is append-only.
· Unknown remains unknown.