EVIDENCE-BOUNDED VALUATION · 2026-07-18

Value what exists. Gate what might.

Advanced decision-intelligence prototype with a functional deterministic public Lab and an external, non-embedded private-product boundary. The analysis separates reusable economic value from a negotiated financing price and from future option value.

Inspect the model Review evidence

Current fair economic / IP value

$0.8–2.0mP50 planning anchor $1.4m

Negotiated financing pre-money

$1.5–4.0mLow-medium confidence · no revenue or customer evidence

Interpretation boundaryIllustrative, evidence-bounded valuation analysis. It is not an appraisal, investment advice, a financing offer, or a prediction.

CONDITIONAL VALUE LADDER

Future value belongs to milestones, not the current balance sheet.

Current economic / IP value$0.80m–$2.0m

Existing reusable code, product design, taxonomy and documentation

Validated MVP$4.0m–$10m

Measured user value, repeat use and benchmarked quality

First paying customer$7.0m–$16m

Paid institutional contract and repeatable delivery

Initial product-market fit$25m–$75m

$2m-$5m ARR, GRR >90%, NRR >105% and repeatable acquisition

TECHNICAL & COMMERCIAL INVENTORY

Observed capability before narrative.

Repository evidence was classified by execution state. Designed or displayed capability is not treated as enterprise product evidence.

Public product

Implemented

Institutional site, deterministic Lab, 6-12 synthetic opportunities, comparison, exports and human review controls.

Public quantitative catalog

Implemented

14 deterministic public models and lenses in code, with explicit assumptions and limitations.

Private agent registry

External boundary

Only aggregate documentation of 1,970 profiles across 51 domains remains public; individual profiles are not embedded or validated here.

Private LLM execution

Not embedded

The public repository fails closed and requires a separately governed private runtime for any specialist execution.

Deployment intelligence

Functional demo

Visible WBS, PERT, critical path, Monte Carlo, cost decomposition and readiness using synthetic planning assumptions.

Enterprise data layer

Planned

No operational multi-tenant mission, evidence, decision or durable audit backend is connected.

Commercial traction

No evidence

No repository evidence of customers, pilots, revenue, ARR, retention, contracts or product-market fit.

Protected IP and calibration

No evidence

No repository evidence of patents, exclusive datasets, historical outcome calibration or validated data flywheel.

PROBABILISTIC DCF · INTERACTIVE

Stress the assumptions before trusting the output.

Each scenario contains a ten-year free-cash-flow path. Change discount, terminal growth and outcome probability; inputs are bounded and never treated as observed facts.

Discount rate 40%Early-stage risk and capital scarcity. Editable assumption.Terminal growth 3%Bounded below the selected discount rate.Scenario probability 15%Subjective probability from the current state, not historical frequency.

Base B2B conditional DCF

$0.32m

Forecast-period PV-$0.38m

Terminal-value PV$0.70m

Positive-value probability contribution$0.05m

Year-10 revenue assumption: $29m. Negative DCFs are possible and are not hidden.

Y10

Revenue FCFF Negative FCFF

MONTE CARLO · 200,000 DRAWS · SEED 240719

A median under $0.5m with a very long right tail.

Risk-adjusted economic option value, not a negotiated financing price. The mean is dominated by a small right tail and should not anchor current value.

P10$0.19m

P25$0.28m

P50$0.45m

P75$0.78m

P90$6.6m

MEAN$4.1m

Not the recommended anchor

VALUATION READINESS SCORE

Commercial evidence is the binding constraint.

30/100

Evidence-based readiness

Product65/100 · weight 15%

Broad working prototype; core enterprise workflow is not complete.

Market25/100 · weight 12%

Plausible wedge, no validated ICP or willingness to pay.

Customers0/100 · weight 12%

No evidenced users, pilots or customers.

Technology60/100 · weight 12%

Deterministic models, bounded execution and tests; limited differentiation.

Security35/100 · weight 10%

Useful request controls; no enterprise identity, RBAC or durable audit.

Team15/100 · weight 10%

No evidenced multidisciplinary operating team or delivery capacity.

Revenue0/100 · weight 10%

No evidenced recurring or services revenue.

External validation15/100 · weight 8%

No benchmark against analyst plus general AI baseline.

IP35/100 · weight 6%

Code, taxonomy and know-how exist; legal exclusivity and calibration do not.

Scalability25/100 · weight 5%

Cloud-compatible surface; no multi-tenancy, metering or operating proof.

STAGE VALUATION · CONDITIONAL PRE-MONEY

Valuation rises only after evidence changes.

P50 and P90 are conditional marks at each achieved stage. The separate probability is the subjective chance of reaching that stage from today.

Idea and architecture100% probability from current state

$0.60m P50

Coherent thesis and defensible product boundary

Public demo100% probability from current state

$1.2m P50

Working reproducible workflow and visible limitations

Authenticated preview - current closest100% probability from current state

$2.2m P50

Bounded preview workflows; the private catalog and executor remain external

Functional MVP75% probability from current state

$3.8m P50

One complete workflow, durable records, evals and usage metering

User-validated MVP50% probability from current state

$6.5m P50

Measured time saved, quality gain, repeat use and WTP

Institutional pilot35% probability from current state

$7.5m P50

Named buyer, paid scope, security review and success metrics

First paying customer25% probability from current state

$10m P50

Signed contract and repeatable delivery without founder-only effort

$100k ARR18% probability from current state

$12m P50

Recurring software revenue separated from services

$1m ARR8% probability from current state

$27m P50

Cohort retention, credible CAC and enterprise reference customers

Initial product-market fit5% probability from current state

$45m P50

$2m-$5m ARR, GRR >90%, NRR >105%, repeatable CAC

$10m ARR2% probability from current state

$90m P50

Efficient growth, strong retention and enterprise controls

International expansion1% probability from current state

$140m P50

Multi-region sales, support, data governance and localization

Enterprise multi-industry platform0.3% probability from current state

$350m P50

$30m-$75m ARR, audited controls and proven cross-vertical repeatability

Logarithmic scale · marker = P90 · values are conditional and must not be multiplied into current value without an explicit probability model.

FOCUS & MONETIZATION

Start with one governed mission for CVC and venture studios.

Turn a thesis into an evidence-backed shortlist and an investment-committee-ready decision record. Multi-industry reach remains an option, not current SAM.

RECOMMENDED WEDGECVC, venture studios and corporate innovation / investment teams

VC-ONLY ILLUSTRATIVE SAM$3.6m-$26.9m annualUnvalidated ICP filter and ACV

3-YEAR CAPACITY SOM$0.45m-$2.0m15-40 customers, not a forecast

01
4-6 week paid mission at $8k-$25k
02
$24k-$60k annual workspace after repeatability
03
$75k-$200k enterprise only after security and integration proof
04
API, white label and marketplace only after stable evals, rights and demand

All ICP filters, ACV, conversion, cycle and retention assumptions are unvalidated and must be replaced with observed funnel data.

BUSINESS-UNIT OPTIONS

Sequence the options. Do not sum them.

Recommended first wedge78

Venture intelligence for CVC and venture studios

Head of CVC, venture studio or corporate innovation

Pricing: $24k-$60k annual workspace after paid pilots
Sales cycle: 2-6 months hypothesis
Margin: Medium-high after workflow standardization
Risk: Medium

Second vertical after proof66

Corporate decision intelligence

Strategy, innovation or capital-allocation leader

Pricing: $75k-$200k enterprise hypothesis
Sales cycle: 4-9+ months hypothesis
Margin: High potential
Risk: Medium-high

Transitional learning model62

Research / consulting as a service

Fund, consultancy or strategy team

Pricing: $8k-$25k scoped mission
Sales cycle: 1-3 months hypothesis
Margin: Medium; expert labor intensive
Risk: Medium

Option after evidence workflow maturity44

University and scientific research intelligence

Research office, lab or technology-transfer team

Pricing: Institutional license; unvalidated
Sales cycle: 6-12 months hypothesis
Margin: Medium
Risk: Medium-high

Product option42

Enterprise simulation and deployment intelligence

Transformation or program executive

Pricing: Enterprise plus implementation; unvalidated
Sales cycle: 6-12+ months hypothesis
Margin: Mixed
Risk: High

Option only31

Financial intelligence

Institutional finance or risk team

Pricing: Unvalidated
Sales cycle: Long
Margin: High potential
Risk: High

Option only; no clinical claims27

Healthcare and life sciences intelligence

Research or evidence leader

Pricing: Unvalidated
Sales cycle: Long
Margin: High potential
Risk: Very high

Option only; safe analytical uses24

Government, defense and resilience

Program, planning or procurement authority

Pricing: Contract-based; unvalidated
Sales cycle: 12-24+ months hypothesis
Margin: Variable
Risk: Very high

Defer until evals, rights and demand20

Agent, model, API and marketplace infrastructure

Developer or platform team

Pricing: Usage / licensing; unvalidated
Sales cycle: Unknown
Margin: High potential
Risk: Very high

DEPLOYMENT INTELLIGENCE

Capital and time to make the option real.

Finance overlays are intentionally wider than the repository's synthetic demo cost distribution.

Target state	Time P50 / P90	Cost P50 / P90	Capital	Evidence gate
Paid design-partner pilot	6 / 9 mo	$0.75m / $1.2m	$0.6m-$1.2m	Paid scope, baseline benchmark and named decision owner
Initial production	12 / 18 mo	$1.8m / $2.8m	$1.5m-$3.0m	Durable records, enterprise identity, security review and supported operations
Initial product-market fit	24 / 36 mo	$3.2m / $5.0m	$2.0m-$5.0m cumulative	Retention, repeatable acquisition, reference customers and measured outcome quality
Enterprise expansion	36 / 54 mo	$6.5m / $10m	$6.0m-$10.0m cumulative	Multi-tenant controls, integrations, observability and repeatable implementation

1
Scope and acceptance criteria
2
Data provenance and integrations
3
Core repeatable workflow
4
Threat model, identity and compliance
5
Paid pilot validation
6
Deployment and operating controls
7
Predicted-versus-actual learning

VALUE-DESTRUCTION RISKS

The biggest risks are commercial proof, differentiation and enterprise controls.

Risk	Probability	Impact	Mitigation	Owner	Budget
No validated willingness to pay	High	Critical	20-30 ICP interviews, three pricing tests and five paid scoped pilots	Founder / commercial lead	$100k
Thin wrapper over general AI	High	Critical	Benchmark against analyst plus general-model baseline; build evidence and outcome workflow	Product and research	$180k
Founder-heavy services prevent scale	High	High	Measure implementation hours and standardize one governed mission	Delivery lead	$120k
Broad multi-industry positioning	High	High	Commit to CVC / venture-studio wedge and explicit option gates	CEO / strategy	$25k
No exclusive data or outcome rights	High	High	License connectors and contract for governed outcome capture where permitted	Data and legal	$250k
Enterprise security gap	High	Critical	Organizations, RBAC, durable audit, secrets, retention, pen test and incident process	Security lead	$450k
Agent catalog quality is unvalidated	High	High	Separate documented, runnable, tested, expert-reviewed and validated statuses	Evaluation lead	$220k
Hallucinated or weak evidence	Medium-high	Critical	Source rights, provenance, retrieval evaluation, abstention and human review	Evidence lead	$300k
Long enterprise sales cycles	High	High	Paid design partners, narrow scope and milestone-based conversions	Commercial lead	$150k
Provider dependency and cost variance	Medium	Medium	Metering, budgets, model abstraction, caching and safe fallback	Engineering	$100k
Regulated-sector overreach	Medium	Critical	Keep health, defense, government and finance as gated options	Governance / legal	$80k
Insufficient team capacity	High	High	Build a 6-8 FTE product, security, domain and commercial team in phases	Founder / board	$900k

COMPETITORS & SUBSTITUTES

General AI plus internal build is the sharpest substitute.

Private-market intelligence

PitchBook; CB Insights; AlphaSense; Tegus / Quid

Incumbent strength: Proprietary data, distribution, workflow and trusted research

Current gap: No comparable data breadth, rights, customer base or distribution

Complement data systems; do not claim to replace them

Enterprise decision and data platforms

Palantir; ServiceNow; Dataiku; Databricks; Snowflake; C3.ai

Incumbent strength: Integrations, ontology, governance, operations and enterprise procurement maturity

Current gap: No multi-tenant control plane, production integrations or service history

Not valuation comparables at current stage

General AI and internal build

Microsoft Copilot; OpenAI API; Google Vertex AI; AWS Bedrock

Incumbent strength: Low marginal model cost, distribution and developer ecosystems

Current gap: Generic synthesis and orchestration are easy to reproduce

Most dangerous substitute

Open-source agent frameworks

LangGraph; AutoGen

Incumbent strength: Durable orchestration and human-in-the-loop primitives

Current gap: Agent count and routing alone are not a moat

Defend with evidence, outcomes, integrations and institutional history

Consultancies and internal analysts

Strategy firms; corporate strategy; data science teams

Incumbent strength: Trust, customization, domain context and accountability

Current gap: Current product still needs expert delivery

Potential design partners and channel, not only competitors

VALUE-CREATION ROADMAP

The next year is an evidence program.

Next 90 days

Choose CVC / venture studios as the only active wedge
Run 20-30 buyer interviews and three explicit pricing tests
Secure five paid scoped design-partner pilots
Benchmark one mission against analyst plus general AI
Measure evidence error, time-to-output, expert hours and compute cost

Next 180 days

Convert at least two pilots to annual contracts
Ship durable mission, evidence, decision and audit records
Implement organizations, RBAC, usage metering and cost budgets
Publish agent / workflow evaluation statuses and versioned model records
Prove a repeatable delivery playbook not dependent on founder-only labor

Next 365 days

Reach $100k-$300k ARR with software and services separated
Show first GRR, NRR, CAC payback and gross-margin cohorts
Complete a security assessment and remediate high-severity findings
Add licensed evidence integrations and governed outcome capture
Decide with data whether corporate strategy is the second vertical

METHODS, ASSUMPTIONS & SOURCES

Auditable inputs, explicit limits.

35% triangulation weight

Cost to recreate - current repo

$0.65m–$1.3m

30-55 person-months; replacement cost is not market value.

15% triangulation weight

Risk-adjusted commercial option DCF

$0.00m–$0.50m

Limited-liability option value; raw probability-weighted project NPV is negative.

20% triangulation weight

Venture capital method

$0.50m–$4.8m

Base pre-money after a $1m raise; highly sensitive to exit, dilution and return hurdle.

15% triangulation weight

Correlation-haircut SOTP

$0.80m–$2.2m

Avoids adding correlated product, method and option values at face value.

15% triangulation weight

Stage / financing reference

$1.5m–$4.0m

Negotiated pre-money range, not intrinsic value.

Unit-economics assumptions

Base: $90k ACV, 72% GM, $45k CAC, 8.3 month payback, 96% NRR.

Downside: $60k ACV, 60% GM, $70k CAC, 23.3 month payback.

Editable planning assumptions only. No observed Venture Governance unit economics exist. Separate software and services margins.

Source register

Observed repository evidenceRepository README · offline repository evidence
Observed repository evidencePlatform architecture and capability labels · offline repository evidence
Observed repository evidencePublic Lab limitations · offline repository evidence
Observed repository evidenceExecution feasibility engine · offline repository evidence
External market evidenceNVCA 2026 Yearbook
Competitor primary sourcePitchBook products
Competitor primary sourceCB Insights Strategy Terminal
Competitor primary sourceAlphaSense enterprise deployment
Public-company primary sourcePalantir 2025 Form 10-K
Competitor primary sourceServiceNow Strategic Portfolio Management
Competitor primary sourceDataiku agent governance release
Provider primary sourceOpenAI GPT-5.4 API pricing
Framework primary sourceLangGraph documentation
Framework primary sourceMicrosoft AutoGen documentation