Hard-to-fool blueprint - hourly run 2026-06-28 22:00 America/New_York

California SB 253 GHG Reporting Readiness Desk

A done-for-you, assurance-ready Scope 1 and Scope 2 emissions reporting packet service for U.S.-based companies with more than $1B in revenue doing business in California.

1. Thesis

Verified: California SB 253 requires large U.S.-based companies doing business in California to disclose Scope 1 and Scope 2 greenhouse gas emissions in the first reporting year, with Scope 3 beginning later. CARB issued a June 2026 update proposing to defer the 2026 Scope 1 and Scope 2 reporting deadline from August 10, 2026 to November 10, 2026 while clarifying the initial regulation. The business sells a finished, reviewer-approved SB 253 readiness packet: entity applicability screen, facility/source inventory, utility and fuel evidence map, GHG Protocol calculation workbook, exception log, public-disclosure draft, and assurance handoff room. It is not a customer-facing co-pilot; customers hand over records and receive a compliance-ready evidence package through a climate-accounting expert interface.

2. Discovery Rationale

The scan covered healthcare quality submissions, FDA food traceability, EPA chemical reporting, DOT compliance records, California climate disclosure, and refrigerant/HFC reporting. SB 253 won because the market has a fresh regulatory event, a clearly defined enterprise buyer, an official deadline update, a public statutory threshold, recognized accounting standards, vendor precedent, and a narrow first-year Scope 1/2 packet that can be tested before Scope 3 complexity arrives.

Rejected candidates lost on urgency, duplicate risk, weaker buyer clarity, or overdependence on legal judgment. FSMA 204 traceability is real but the compliance date is farther out. Hospice/SNF quality reporting was too adjacent to prior PBJ staffing. EPA TSCA/PFAS and CLFS were already covered by prior blueprints. DOT DBE/compliance records had less fresh buyer-pull evidence in this run.

3. Candidate Comparison

Candidate	Buyer	Outcome	Score summary	Evidence quality	Dynasty type	Decision
California SB 253 GHG Reporting Readiness Desk	Large-company CFO, sustainability controller, ESG reporting lead, legal ops	Assurance-ready Scope 1/2 packet and disclosure workpaper	High urgency, high document burden, clear deadline, strong AI leverage; judgment contained by expert review	Verified deadline, scope, thresholds, standards; Inferred willingness to pay	Dynasty vertical blueprint	Selected
FSMA 204 Traceability Readiness Desk	Food manufacturer, distributor, restaurant chain supply-chain lead	KDE/CTE traceability map and FDA 24-hour response drill	Strong workflow fit but longer runway after deadline extension	Verified FDA rule and extension; buyer urgency inferred	Dynasty vertical blueprint	Rejected this hour: less time pressure
Hospice CAHPS Data Submission Desk	Hospice quality leader, compliance officer	Quarterly CAHPS submission readiness and exception packet	Repeatable, but narrow and close to prior healthcare quality reporting outputs	Verified CMS deadlines; pricing unverified	Dynasty module bundle	Rejected: duplicate-adjacent
HFC Technology Transitions Reporting Desk	Manufacturers/importers using HFCs	Annual EPA reporting packet	Clear EPA reporting, but deadline may have passed for first cycle	Verified EPA forms and reporting; urgency inferred	Dynasty module bundle	Rejected: timing weaker
DOT DBE Documentation Readiness Desk	DOT-funded prime contractor or recipient compliance leader	DBE documentation and audit-response binder	Document-heavy, but regulatory trigger and paid demand were weaker in this run	Verified DOT guidance; demand unverified	Dynasty capability expansion	Rejected: demand evidence gap
EPA TSCA Section 8(d) Health and Safety Data Desk	Chemical manufacturer/importer EHS lead	CDX-ready health and safety study submission packet	Good fit but overlaps recent PFAS/TSCA blueprint shape	Verified EPA extension; inferred packet demand	Dynasty vertical blueprint	Rejected: duplication risk

4. Hard Disqualifier Check

Disqualifier	Status	Reason
Customer-facing co-pilot or SaaS	Pass	Service sells a finished evidence packet and disclosure workpaper, not software access.
Physical labor requirement	Pass	Records, utility bills, fuel logs, leases, invoices, ERP exports, and facility metadata can be handled remotely.
Hourly or cost-plus pricing dependency	Pass	Priced per entity, facility band, and reporting packet, with optional annual refresh.
Cannot plausibly reach 50%+ gross margin	Pass	Initial packets are expert-heavy; mature repeats should automate extraction, mapping, variance checks, and workpaper generation.
Buyer unclear	Pass	Economic buyer is CFO/controller or legal/compliance executive; operational owner is sustainability reporting lead.
Workflow not decomposable	Pass	Applicability, boundary setting, source inventory, evidence intake, calculations, QA, review, and disclosure are separable.
Improper automation of regulated judgment	Pass	Climate-accounting expert approves methodology and public claim boundaries; assurance providers stay independent.
Duplicative of prior blueprint or Dynasty	Pass	Different buyer, climate data inputs, regulatory trigger, and assurance-readiness workflow.
Illegal or impossible licensing model	Pass	Service avoids legal advice and third-party assurance opinions; coordinates with counsel/assurance providers.
Core demand unverified	Pass	Official CARB rulemaking and deadline updates verify compliance trigger; paid demand remains inferred.
Commoditized by frontier AI	Unclear	Generic AI may draft calculations, but defensibility comes from evidence provenance, review network, and repeatable controls.
Cannot test with bounded pilot	Pass	First pilot can cover one entity and 5-15 facilities for Scope 1/2 only.

5. Rubric Scorecard

4/5Low trust burden

Large companies already buy climate accounting, ESG advisory, and assurance readiness from external firms.

4/5Low task-level judgment

Most work is evidence collection, normalization, mapping, and calculation checks; methodology exceptions need expert review.

4/5High intelligence threshold

Requires synthesis across entities, facilities, invoices, emission factors, accounting boundaries, and disclosure text.

4/5Regulation as moat

Regulatory deadline, GHG Protocol alignment, public reporting, and future assurance create trust and process barriers.

5/5No physical labor

Delivery is remote and document/data-based.

4/5Sam Altman test

Better models improve extraction, reconciliation, source citation, and anomaly detection, but expert governance remains key.

6. Opportunity

>$1BVerified SB 253 reporting-entity revenue threshold for U.S.-based companies doing business in California.

Nov. 10, 2026Verified CARB's proposed deferred 2026 Scope 1/2 reporting deadline.

Scope 3 in 2027Verified SB 253 first year covers Scope 1/2; later years include Scope 3.

2,596Verified CARB workshop estimate of SB 253 entities used for fee assumptions.

$3,106Verified CARB workshop estimated annual SB 253 entity fee, adjusted over time.

50-60%Inferred target mature gross margin after repeatable intake, evidence mapping, and expert review reuse.

7. Evidence Quality And Source-Claim Matrix

Claim	Label	Source or basis	Confidence	Business impact
SB 253 requires U.S.-based companies with more than $1B revenue doing business in California to disclose GHG emissions.	Verified	CARB June 2026 update and CARB February 2026 news release	High	Defines ICP and trigger.
CARB is proposing to defer the 2026 Scope 1/2 deadline from August 10 to November 10, 2026.	Verified	CARB GovDelivery notice, June 2026	High	Creates timely remediation window.
First-year SB 253 reporting covers Scope 1 and Scope 2, with Scope 3 beginning in subsequent years.	Verified	CARB GovDelivery and CARB news release	High	Allows a bounded first product.
CARB estimated 2,596 SB 253 entities and $3,106 annual fee in public workshop materials.	Verified	CARB public workshop slide result surfaced in search	Medium	Signals market size and administrative seriousness.
GHG Protocol provides standards and Scope 2 guidance for corporate emissions accounting.	Verified	GHG Protocol standards pages	High	Defines knowledge base and calculation rules.
Companies already buy SB 253 compliance software and assurance-readiness services.	Verified	Vendor pages from Watershed, Sinai, Certivo, Verdis, Johnson Lambert	Medium	Confirms vendor precedent, not direct willingness to pay for this service.
Best wedge is a done-for-you packet rather than another carbon accounting platform.	Inferred	Derived from vendor crowding plus customer data-normalization burden	Medium	Positions against SaaS fatigue.
Target gross margin can exceed 50% after 20-30 packets.	Inferred	Modeled from repeatable extraction, facility templates, and reviewer minutes	Low	Must be proven in pilots.
Enterprises will pay $18k-$75k for a packet depending on facility complexity.	Unverified	Plausible enterprise compliance-services pricing; no direct public quote found	Low	Needs paid discovery calls.

8. Why Now

Verified market/regulatory changes

CARB approved initial regulation in February 2026, submitted a package to OAL in May 2026, and issued a June 2026 update proposing a three-month reporting deadline deferral to November 10, 2026.

Inferred AI capability changes

Modern models can extract facility names, meter IDs, fuel quantities, dates, tariff classes, invoice totals, leases, and methodology notes from messy PDFs and spreadsheets, then reconcile them against deterministic rules.

Unverified hypotheses

Some in-scope companies may still lack centralized Scope 1/2 evidence and will prefer a managed packet over a new platform implementation. This must be validated through paid pilots.

9. Customer & PMF

ICP	U.S.-based companies over $1B revenue doing business in California, especially multi-site retail, logistics, food, healthcare, hospitality, manufacturing, and consumer services.
Buyer	CFO, controller, general counsel, chief sustainability officer, audit committee sponsor.
Economic decision-maker	Finance/legal executive accountable for public disclosure, risk, and audit readiness.
User	ESG reporting manager, energy manager, facilities operations, procurement analyst, external assurance team.
Urgent trigger	CARB 2026 Scope 1/2 reporting deadline and board pressure to avoid public disclosure errors.
Alternatives	Big 4 advisory, carbon accounting SaaS, ESG consultants, internal spreadsheets, do nothing until final CARB forms stabilize.
Jobs-to-be-Done	"Get my entity boundary, source data, calculations, exceptions, and disclosure narrative into an audit-ready state before the reporting date."
Willingness-to-pay evidence	Inferred: vendor market exists and buyer is enterprise compliance. Direct price validation remains unresolved.

10. The Outcome We Sell

Paid outcome: a reviewer-approved SB 253 Scope 1/2 readiness packet for one reporting entity.

Deliverable: applicability memo, entity/facility boundary register, source inventory, evidence index, calculation workbook, exception log, public disclosure draft, and assurance-provider handoff room.
Acceptance criteria: every emissions number traces to a source record, calculation method, emission factor, reviewer note, and exception disposition.
Customer promise: the packet is ready for executive review, counsel review, and assurance-provider scoping. It does not promise CARB acceptance or assurance opinion.
Exclusions: legal advice, independent assurance, final signatory responsibility, Scope 3 in the first package unless separately contracted.
Rework policy: fix packet defects caused by our mapping/calculation errors at no charge within 30 days.

11. Internal AI Engine Architecture

Intake

Secure upload of utility bills, fuel invoices, fleet records, leases, facility lists, ERP exports, meter reports, prior ESG reports, and organizational charts.

Normalization

Entity, facility, meter, source, date, unit, supplier, and document version are normalized into governed tenant objects.

Retrieval and knowledge

GHG Protocol Corporate Standard, Scope 2 Guidance, EPA Scope 1/2 guidance, CARB notices, customer policy, and prior approved packet examples.

AI workbench

Extracts records, classifies source types, drafts exception explanations, reconciles inconsistent names, and prepares reviewer-ready summaries.

Deterministic rules

Unit conversion, source-category mapping, date coverage, duplicate bill detection, boundary checks, and formula validation are rules-owned.

Human chokepoints

Climate accountant approves boundary, methodology, factors, exceptions, market/location Scope 2 treatment, and public narrative.

QA and delivery

Second reviewer checks traceability, variance, missing-source flags, and disclosure consistency before customer delivery.

Learning loop and portability

Reviewer edits become test cases and gold examples; extraction prompts and validators are model-portable.

12. AI-vs-Human Operations Pipeline

AI-owned
Document OCR, field extraction, facility/entity matching, draft source inventory, invoice anomaly summary.

Rules-owned
Unit conversions, formula checks, duplicate detection, period coverage, required field completeness.

Operator-reviewed
Missing records, unclear meters, unmatched suppliers, customer follow-up queue.

Expert-approved
Boundary, methodology, emissions factors, exception disposition, disclosure language.

Customer-facing
Kickoff, evidence request, weekly exception report, final packet walkthrough.

13. Operations As Product

The operating system is the product. Every packet runs through a fixed launch checklist: entity applicability screen, facility/source inventory, evidence request, completeness score, extraction batch, deterministic validations, expert review, red-team check, customer exception sprint, and final packet lock. Variance is reduced through source-specific SOPs, required evidence lists, confidence scoring, reviewer assignment logic, audit trails, version control, gold-standard packets, root-cause analysis on rejected lines, and postmortems for every customer rework request.

14. No-Holes Quality Engine

Every calculated number must link to a source record, period, unit, factor, and formula.
AI outputs cannot write final numbers without deterministic recalculation.
Hallucination guard: source-only generation and mandatory citation back to uploaded evidence.
Completeness guard: facility list, source categories, and date coverage must reconcile before final delivery.
Disclosure guard: public language is reviewed against counsel-approved claim boundaries.
Assurance guard: packet separates prepared-by evidence from independent assurance opinion.

15. Pricing, Pricing Legality, And Unit Economics

Primary pricing	Fixed per reporting entity: $18k starter for up to 10 facilities, $35k core up to 50 facilities, $75k+ enterprise multi-entity packet. Annual refresh at 45-60% of first-year price. Price levels unverified.
Why not hourly	Hourly billing rewards messy inputs and expert labor. Packet pricing rewards repeatability and automation.
Pricing legality	Fixed-fee operational readiness is preferable. Avoid contingency pricing tied to regulatory outcome, penalty avoidance, or assurance opinion.
COGS per core packet	Project lead $3,000; climate accountant $4,500; operator $2,000; QA reviewer $1,250; model/OCR/storage $250-$700; secure workspace $200-$600; total target $11k-$13k after maturity.
Model cost per unit	Inferred: $250-$700 for OCR, extraction, summarization, and validation passes depending on document volume.
Human review minutes	Launch target 900-1,500 minutes per core packet; mature target 360-600 minutes as source templates and validators improve.
Licensed review cost	No licensed legal opinion included. Climate-accounting expert review is required; counsel and assurance provider remain external if needed.
Gross margin target	First 10 packets: 25-40%. Mature repeatable packets: 50-60%.
Revenue per FTE target	$450k launch target; $800k+ mature operator pod target.

16. Nonlinear Scaling Plan

Start with small facility bands and one fiscal year. By packet 20, reusable source parsers should cover utility bills, fuel cards, refrigerant logs, fleet exports, and facility master lists. Mature automation target is 65-75% of line-level evidence extraction and validation, with 10-20% expert escalation. Revenue decouples from headcount as the same expert supervises more packets, reviews exceptions instead of raw records, and turns corrections into validators.

17. Moat & Sam Altman Test

Frontier model improvement strengthens the company if the proprietary asset is not prompts but a governed evidence factory: source-linked workpapers, accepted reviewer patterns, facility/source templates, variance libraries, and assurance-room packaging. Better models lower extraction and reconciliation costs while increasing throughput. The strongest commoditization threat is carbon-accounting platforms adding managed services; the response is to own the high-trust packet handoff, expert network, and deadline-driven implementation sprint.

18. Buyer-Specific Go-To-Market

Motion: founder-led outbound plus climate-accounting consultants, mid-market assurance firms, energy brokers, and ESG counsel channels.
First 50 prospects: multi-site companies likely over $1B revenue in retail, logistics, hospitality, food distribution, healthcare services, manufacturing, and facilities-heavy consumer services.
Trigger events: CARB deadline update, board/audit committee ESG agenda, assurance provider scoping, prior public sustainability report gaps.
Outreach wedge: "We turn your Scope 1/2 evidence into an assurance-ready SB 253 packet in four weeks without making your team operate another platform."
Credibility asset: redacted sample packet with source-to-number traceability and expert reviewer checklist.
Conversion path: 30-minute applicability screen, paid $4,500 mini-audit, then fixed-fee packet.
Sales cycle: 3-8 weeks depending on legal/procurement and data-security review.
Metrics: paid mini-audit conversion, evidence upload rate, exception closure time, packet acceptance.

19. Pilot Design And Early-Demand Trap Mitigation

Cap the pilot at three customers, one entity each, and no more than 15 facilities per customer. Success means customer accepts the source-linked packet, external reviewer finds no material calculation defect in a sample, and operator time falls below 45 minutes per evidence line by the final batch. Track every manual workaround. Kill or narrow if customers refuse paid mini-audits, data access takes over three weeks, or expert review finds methodology ambiguity in more than 20% of lines.

20. Competitive Landscape

Category	Examples	How to compete
Carbon accounting SaaS	Watershed, Persefoni, Sinai, Sweep, Certivo, Optera	Do the work for teams that cannot implement another platform before the deadline.
Consulting and assurance firms	Big 4, ESG boutiques, Johnson Lambert, climate consultants	Partner when they need evidence prep; do not compete for formal assurance opinion.
Internal teams	Finance, ESG, energy management, facilities	Replace spreadsheet chase work and create a consistent packet.
Do nothing	Wait for final CARB forms or rely on good-faith discretion	Sell readiness, board comfort, and faster correction of evidence gaps.

21. Regulation, Compliance, And Licensing Boundary

The service prepares operational evidence and workpapers. It does not provide legal advice, does not act as independent assurance provider, and does not certify emissions. Counsel approves customer-specific legal claims and public-disclosure language where required. Assurance providers receive the packet but remain independent. Privacy controls include role-based access, source-record retention policy, audit logs, encrypted storage, customer-approved deletion, and no confidential material in public submissions without approval.

22. Compact Founding Team And Expert Map

Role	Why needed	Timing
Climate-accounting lead	Approves GHG Protocol methodology, factor selection, and exception logic.	In-house or deeply retained from day one.
Operations lead	Runs intake, customer follow-up, packet QA, and delivery cadence.	In-house from day one.
Automation/AI engineer	Builds extraction, normalization, validator, and packet renderer.	Fractional until pilot closes.
Compliance counsel	Reviews engagement terms, prohibited claims, and disclosure boundary.	Fractional.
Channel/sales lead	Gets assurance, ESG, counsel, and energy broker partners.	Founder-led first 90 days.
QA owner	Maintains gold packets, red-team checks, and rework postmortems.	Operations lead initially.

23. Exhaustive Risk Register

1. CARB deadline or requirements change again

Likelihood: Medium. Impact: High. Label: Verified regulatory volatility. Mitigation: deadline monitor, change log, modular packet templates. Owner: compliance lead. Indicator: new CARB notices.

2. Buyer waits due to good-faith enforcement discretion

Likelihood: Medium. Impact: High. Label: Inferred. Mitigation: sell board/audit readiness and Scope 3 runway, not panic. Owner: founder. Indicator: paid mini-audit conversion below 20%.

3. Data access stalls inside enterprise

Likelihood: High. Impact: High. Label: Inferred. Mitigation: start with facility source inventory and executive-owned evidence request. Owner: operations lead. Indicator: first upload later than day 7.

4. Methodology disputes create too much expert labor

Likelihood: Medium. Impact: High. Label: Inferred. Mitigation: narrow to Scope 1/2, use counsel-approved disclaimers, escalate only exceptions. Owner: climate-accounting lead. Indicator: expert minutes exceed 1,500.

5. SaaS platforms bundle managed services

Likelihood: High. Impact: Medium. Label: Verified vendor precedent. Mitigation: partner with platforms and specialize in packet rescue. Owner: sales lead. Indicator: vendor service pages promise same outcome.

6. Assurance providers reject packet format

Likelihood: Medium. Impact: High. Label: Unverified. Mitigation: interview assurance teams before pilot, map packet to their request lists. Owner: QA owner. Indicator: more than 10% sample rework.

7. Public disclosure creates legal exposure

Likelihood: Medium. Impact: High. Label: Verified reporting is public. Mitigation: counsel review, no certification claims, customer signoff. Owner: compliance counsel. Indicator: unresolved disclosure comments.

8. Security review blocks enterprise sales

Likelihood: Medium. Impact: Medium. Label: Inferred. Mitigation: encrypted storage, DPA, SOC 2 roadmap, customer-hosted option. Owner: operations lead. Indicator: security review over 14 days.

9. Margin fails due to messy facility records

Likelihood: Medium. Impact: High. Label: Inferred. Mitigation: facility caps, evidence quality surcharge, paid mini-audit. Owner: founder. Indicator: COGS above 55% after packet 10.

10. AI extraction errors damage trust

Likelihood: Medium. Impact: High. Label: Inferred. Mitigation: deterministic recalculation, source citations, two-person QA on material lines. Owner: QA owner. Indicator: material QA defect rate over 1%.

11. Scope 3 expansion is too complex

Likelihood: High. Impact: Medium. Label: Verified future Scope 3 requirement. Mitigation: keep first offer Scope 1/2 and design supplier evidence module later. Owner: product lead. Indicator: customers demand Scope 3 before core repeatability.

12. No direct willingness to pay

Likelihood: Medium. Impact: High. Label: Unverified. Mitigation: require paid mini-audits before building generalized platform. Owner: founder. Indicator: high meeting interest but no paid conversion.

24. Tech Stack & Build Plan

Use a secure web intake, object storage for source documents, Postgres for tenant objects, a document parser/OCR layer, an LLM extraction workbench, deterministic validators in TypeScript, a calculation workbook renderer, a packet renderer, and immutable audit logs. Build sequence: entity/facility model, evidence upload, source parser, Scope 1/2 source classifier, calculation validator, exception queue, expert review screen, customer packet export, assurance-room export, monitoring dashboard.

25. Dynasty Translation Layer

Classification: Dynasty vertical blueprint.

Buyer/outcome fit: clear enterprise buyer and accepted packet outcome.
Module map: applicability screen, facility inventory, evidence intake, source classifier, factor registry, calculation validator, exception queue, expert approval, packet renderer, deadline monitor.
Activation path: guided. Customer configures entity, facilities, fiscal year, source systems, and evidence owners; internal team runs the packet factory.
Tenant data objects: reporting entity, facility, meter, source record, emission source, factor, calculation line, exception, reviewer decision, disclosure draft, packet version.
Warranted claims: readiness packet, source traceability, expert-reviewed workpaper. Unwarranted: CARB acceptance, legal compliance guarantee, independent assurance opinion.
Deployability status guess: spec. Move to implemented after a working intake, parser, validator, and packet renderer process one pilot.
Anti-duplication: not Dynasty Launcher, Your Deputy, HVAC automation, or generic module marketplace; it is a regulated climate-reporting evidence factory.
Recommendation: pilot, with strict paid mini-audit validation before platformization.

26. Metrics & KPIs

Metric	Launch target	Mature target
Packet throughput	2 per month per pod	8+ per month per pod
Cycle time	4-6 weeks	10-15 business days for refreshes
Rework rate	<10% of lines	<3% of lines
Gross margin	25-40%	50-60%
COGS per core packet	$18k-$25k	$11k-$13k
Escalation rate	25%	10-15%
Automation rate	35%	65-75%
Evidence completeness	85%	95%+
Quality failure rate	<2%	<0.5%
Customer acceptance	80%	95%
Pilot conversion	3 paid mini-audits from 30 prospects	40% mini-audit to packet
Dynasty deployability	spec	validated after 3 accepted packets

27. What Could Kill This

The business dies fast if enterprises treat 2026 as a low-risk good-faith year and refuse paid help, if data access is so slow that a packet cannot be delivered inside the deadline window, if assurance firms reject the workpaper format, or if carbon-accounting SaaS vendors absorb the managed evidence-prep job at lower cost. The decisive test is not whether buyers are interested; it is whether they pay for a mini-audit before the final CARB forms are settled.

28. 90-Day Validation And Launch Plan

Weeks	Actions	Kill criteria
1-2	Interview 15 ESG/controllers/assurance contacts; collect request lists; finalize redacted packet template.	Fewer than 5 agree pain is active in 2026.
3-4	Sell three $4,500 mini-audits; run facility/source inventory and evidence-gap score.	No paid mini-audits from 30 targeted prospects.
5-6	Build parser and deterministic validator for top five source types; process pilot evidence.	Extraction QA requires full manual redo.
7-8	Deliver first packet; ask external reviewer to sample lines and packet structure.	Material calculation defects or assurance-provider rejection.
9-10	Convert mini-audits to full packets; instrument expert minutes and rework drivers.	COGS above 70% with no automation path.
11-12	Codify Dynasty module specs, packet renderer, and partner playbook; decide pilot/park/build.	No repeatable buyer segment or channel path.