Building Emissions Compliance Engine

A done-for-you managed service that keeps large commercial and multifamily buildings continuously compliant with U.S. Building Performance Standards (BPS) — NYC Local Law 97, Boston BERDO, Washington DC BEPS, and 40+ peer laws — by producing the mandatory annual, professional-certified emissions filing and a penalty-avoidance decarbonization pathway. AI is the production engine; a licensed Registered Design Professional (PE/RA) is the customer-facing trust interface and signatory.

The one sharp insight: BPS laws are a rare compliance category that is both deadline-bound and recurring annually forever, with an escalating, quantified dollar penalty that gets ~50% worse in 2030 — and the mandatory filing legally requires a licensed professional's signature. That combination (annual recurrence + hard money penalty + licensing moat + document/data-heavy work) is the ideal substrate for an AI-native, per-building managed service.

Thesis

Every year, a growing universe of hundreds of thousands of large U.S. buildings must (1) benchmark energy/water use in ENERGY STAR Portfolio Manager, (2) calculate greenhouse-gas emissions against a legally binding, tightening cap, and (3) file a professional-certified compliance report — or pay per-ton or per-square-foot fines. The work is repetitive, data-heavy, rules-based, and unforgiving of error, yet today it is delivered by fragmented local energy-engineering firms billing hourly, and by benchmarking clerks who do data entry but not the emissions/pathway analysis.

An AI-native engine can ingest utility data, building attributes, and prior filings; compute emissions vs. each jurisdiction's limit; draft the certified report and a ranked decarbonization pathway; and route only genuine exceptions and the final sign-off to a licensed PE/RA. This decouples revenue from headcount, prices per building (annual recurring) rather than per hour, and gets stronger every year as more jurisdictions adopt BPS and limits tighten. The moat is the professional signature plus a proprietary, continuously-improving rules library of jurisdiction-specific emissions factors, limits, and filing formats.

Discovery rationale

This run searched across regulated-compliance terrain (real-estate, energy/utility, environmental, tax, healthcare). BPS emerged because it uniquely combines: an active enforcement moment (NYC LL97's first compliance year was 2024, with the first certified reports due May 1, 2025 [S1][S9]); a quantified and escalating penalty ($268/tCO₂e; ~50% tighter limits in 2030 [S1][S4]); rapid multi-jurisdiction expansion (Evanston IL became the first new BPS adopter of 2025; 40–50+ jurisdictions now [S6][S7]); a legally mandated licensed-professional signature (regulation-as-moat [S1]); and a fragmented incumbent base of hourly engineering firms. Adjacent blueprints in the manifest (California climate disclosure SB 253, EPA environmental reporting, EPR packaging, CBAM, EUDR) address corporate or product-level emissions — none address building-level municipal BPS, confirming a genuine whitespace.

Candidate comparison

Five candidates were generated and scored (1–5 across the 20-factor screen; composite shown). The BPS engine won on active demand, recurring revenue, and licensing moat.

CODE validation

Rubric scorecard — six gates

Composite gate score: 4.5 / 5. Clears the evidence threshold.

Target buyer & Jobs-to-be-Done

Jobs-to-be-Done

• "File a correct, PE-certified emissions report on time so I never get a penalty or a violation on my building record."
• "Tell me exactly how far over the limit I am and the cheapest path to get under it before 2030."
• "Handle it across all my buildings and all my cities so I don't manage ten vendors."

The painful problem

A covered building that misses or fails its filing faces two stacked penalties: a non-filing fine (NYC: $0.50/sq ft/month; LL84 benchmarking: $500/quarter up to $2,000/yr [S8][S16]) and an emissions-overage fine ($268/tCO₂e in NYC; $234/tCO₂e Alternative Compliance Payment in Boston [S1][S5]). Because limits tighten ~50% in 2030, a 300,000 sq ft office projected to pay ~$14,500/yr in 2024–2029 jumps to ~$330,000/yr from 2030 — a 22× increase [S3]. Roughly 57% of NYC covered properties already exceed the 2030 caps [S4]. Owners must act, the math is unforgiving, and the required signature makes DIY impractical.

The outcome we sell

Not software, not a dashboard the owner operates. We sell: "Your building is filed, certified, and penalty-safe — and here is your funded path to stay that way through 2030." Concretely, per building per year the customer receives a PE/RA-certified LL97 (or peer-jurisdiction) emissions report filed to the agency portal, the LL84/benchmarking submission, a plain-English compliance status ("under/over limit by X tCO₂e = $Y exposure"), and a ranked, cost-modeled decarbonization pathway with incentive/rebate capture. The customer's experience is an expert who makes the problem disappear.

First one-feature MVP wedge

Evidence summary & claim table

Source-claim matrix

Market & demand evidence · active buyer conversations

Demand is visible in spending, not just interest. NYC has a dense marketplace of firms selling LL97/LL84 compliance (Bright Power, Bright Energy Services, Greenwich Energy Solutions, Energo, EN-POWER, RAND Engineering, UtiliSave), each with dedicated landing pages and quote flows [S13-adj][S16]. Benchmarking-compliance is transacted at published flat per-building fees ($449–$1,000) [S13][S14]. Trade associations (REBNY), law firms (Greenberg Traurig, Phillips Lytle), and management companies (FirstService) publish urgent owner guidance and deadline calendars [S3][S9]. Boards and managing agents actively seek help because directors fear personal fiduciary exposure. On the supply side, the recurring, deadline-driven nature guarantees a fresh buying moment every single year.

Competitive landscape & budget validation

Budget already exists and is redirectable: owners currently pay hourly engineers, benchmarking vendors, and (worst case) fines. Our per-building recurring fee replaces fragmented spend with one predictable line item — and is trivially justified against a single avoided penalty.

Pricing evidence & proposed pricing · unit economics

Pricing is per building / per filing (recurring annual) plus outcome-linked add-ons — never hourly. Anchored to observed flat benchmarking fees ($449–$1,000/building [S13][S14]) and to the value of a single avoided penalty (tens to hundreds of thousands of dollars [S3]).

Illustrative unit economics (per emissions filing at maturity)

*Launch margin understates true unit economics because early PE minutes double as SOP/rules-library creation. Automation % of steps: ~55% launch → ~75% day 90 → ~85% year 1. Target revenue/FTE: $500k+. CAC payback target <6 months given annual recurrence.

Regulatory & compliance considerations · licensing boundary

Key frameworks: NYC Admin Code §28-320 (LL97) and §28-309 (LL84); Boston BERDO (Ch. 7 Env. Ordinance); DC Clean Energy DC Act / BEPS; state statutes (Colorado HB21-1286, Washington HB1257). Professional engineering licensure governed by state PE boards. Data handling: utility data is commercially sensitive but not typically PII; standard confidentiality + SOC 2 posture over time.

AI-native advantage

AI changes the economics, not just the tooling. Each building's data (utility bills in many formats, Portfolio Manager exports, building attributes, prior filings) is extracted and normalized by AI; jurisdiction-specific emissions factors and limits are applied by a deterministic rules engine; the certified report and pathway are drafted by AI; only true exceptions and the final signature touch a human. As frontier models improve, extraction accuracy rises, more jurisdictions are added faster, and per-unit PE minutes fall — margin expands while the licensed signature keeps the moat intact. The customer never operates AI; they receive an outcome.

• AI tasks: bill/document extraction, unit reconciliation, emissions computation drafting, pathway modeling, report narrative, format-to-portal mapping.
• Deterministic rule tasks: jurisdiction limits, emissions coefficients, exemption logic, deadline calendars.
• Human tasks: exception adjudication, edge-case judgment, PE/RA certification & filing, client relationship.
• Never fully automated: the professional certification and any dispute/legal interaction.

Internal AI engine architecture

AI-vs-human operations pipeline

Dynasty translation layer

• Buyer translation: "Building owner who will be fined unless a certified emissions report is filed" → offer: "We file it, certify it, and keep you penalty-safe."
• Service translation: Done-for-you filing + pathway. Automation handles data + drafting; humans handle exceptions + certification.
• Workflow translation: Intake → normalize → compute → draft → review → certify → file → monitor → renew.
• Tooling translation: OCR + LLM API + Portfolio Manager API + a rules library (Postgres/JSON) + a filing checklist + a CRM. Custom software comes later.
• Sales translation: "Your building is on the covered list. Your report is due May 1. One avoided penalty pays for a decade of our service."
• Delivery translation: First 3 buildings done semi-manually with a partner PE; automate extraction and drafting after patterns stabilize.
• Expansion translation: Per-city rules packs → portfolio subscriptions → retrofit-incentive capture → software-assisted self-serve for the smallest buildings.

Anti-duplication analysis

Existing: decarbonization SaaS (owner-operated dashboards), hourly energy-engineering firms, benchmarking data-entry clerks, enterprise ESG consultancies. Why this is not a copy: we sell a done-for-you, PE-certified outcome priced per building/recurring — not a tool the owner runs, not hourly consulting, not mere data entry. Narrow wedge: the mandatory certified LL97-class filing for mid-market single-building owners and boards, the segment enterprise firms ignore and clerks can't certify. Under-served segment: owners of one to a few buildings, co-op/condo boards, and multi-city portfolios with no single vendor covering all jurisdictions. Unmet pain: the emissions math + pathway + certification + multi-jurisdiction coverage, delivered as one predictable annual product. Distinct from manifest neighbors (SB 253 corporate climate disclosure, EPA reporting, EPR, CBAM, EUDR), which are corporate/product-level, not building-level municipal BPS.

Anti-commoditization analysis (Sam Altman test)

If a future general model lets an owner self-generate an emissions estimate, they still cannot self-certify — the filing legally requires an RDP signature under professional liability [S1]. Our defensibility compounds through: (1) the licensed-signature moat; (2) a proprietary, continuously-versioned rules library across 40+ jurisdictions that is expensive to assemble and keep current; (3) accumulated gold-standard filings and QA checks; (4) portal integrations and audit trails agencies trust; (5) switching costs from year-over-year continuity. Better models make our production cheaper and faster (margin up), not our service obsolete. The service gets stronger as models improve and as more jurisdictions adopt BPS.

Operations as product · no-holes quality engine

• SOPs per jurisdiction; structured intake checklists; required-evidence lists (12 mo utility data, GFA proof, occupancy).
• Automated completeness checks block filing until inputs are whole.
• Exception queue with reviewer assignment; confidence scoring on every computed value.
• Deterministic math in code (not model-estimated); prior-year cross-check; red-team plausibility test.
• Gold-standard example filings per jurisdiction; version control on rules library and templates.
• Full audit trail (inputs, factors used, who certified, timestamp) for agency defensibility.
• Root-cause + postmortem loop on any rejected or corrected filing → new SOP/rule/QA check.

What the human expert actually does

Audit trail required for every certified filing: inputs, emissions factors, computed result, exceptions, certifying RDP, timestamp.

Minimum viable offer · fulfillment · tools

First paid offer: "LL97 Emissions Report, PE-Certified & Filed — $2,000/building, on-time guarantee." First 3 customers fulfilled semi-manually: owner shares utility bills → analyst computes in a spreadsheet cross-checked by the rules engine → partner PE certifies & files via BEAM. Day-one tools: ENERGY STAR Portfolio Manager, OCR (Textract/Document AI), an LLM API, a jurisdiction rules table, a CRM (HubSpot), e-sign, and the DOB BEAM portal. Automate later: extraction, drafting, monitoring, multi-city rules packs. Do not automate first: certification, exception judgment.

Nonlinear scaling & build-before-scale checkpoints

• Revenue/FTE target $500k+; gross margin 55% launch → 78%+ year 1.
• Throughput target: 6–8 filings/analyst/day at maturity; cycle time <5 business days.
• Rework target <5%; quality-failure (rejected filing) <1%; escalation <8%.
• After 5 pilots: harden intake, evidence requirements, QA checks.
• After 10 pilots: harden SOPs, exception queue, reviewer checklists, delivery templates.
• After 20 pilots: pause new intake until COGS, rework, escalation, and cycle time are measured and within target.

Distribution proof table

First 30 days of content

10 educational posts

• "Is your building on the LL97 covered list? How to check your BBL in 2 minutes."
• "$268 a ton: exactly how the LL97 penalty is calculated."
• "Why 2030 is the year LL97 fines get ~50% worse."
• "LL84 vs LL97: the two filings owners confuse (and the fines for each)."
• "Who is legally allowed to certify your emissions report?"
• "The May 1 deadline and the grace period, explained."
• "Co-op boards: your personal fiduciary exposure under LL97."
• "Boston BERDO's $234/ton ACP vs NYC's $268 — a portfolio owner's guide."
• "5 data mistakes that get an emissions filing rejected."
• "The cheapest first move to close your 2030 emissions gap."

3 diagnostic teardown formats

• Live "penalty exposure teardown" of an anonymized building.
• Before/after: a rejected filing fixed.
• "Portfolio heat-map": which of your buildings breach 2030 limits.

2 lead-magnet angles

• Free LL97 Penalty Exposure Estimate (enter address + utility totals).
• Free "Am I covered?" BBL lookup + deadline calendar.

1 webinar  ·  1 outbound diagnosis template

Webinar: "Avoiding LL97 fines: a board & owner playbook for this filing season." Outbound: "We ran your building (BBL __) against the LL97 2024–2029 and 2030 limits — you're ~X tCO₂e over, ≈$Y/yr exposure. Want the certified filing handled?"

Lead magnet, waitlist & warm GTM

Lead magnet / value exchange: a free LL97 Penalty Exposure Estimate — the buyer enters address and utility totals and receives an instant, credible exposure number and deadline. This captures the exact pain signal (over-limit buildings) and qualifies leads by dollar exposure. Waitlist: "Reserve this filing season's slot" cap to create urgency and control pilot intake. Warm GTM: follow up estimate users and webinar signups with a scoped consult; convert boards via their managing agents. Qualification: covered building + measurable over-limit exposure + decision-maker identified = sales-ready.

Pilot design & early-access feedback flywheel

• First cohort: 8–10 NYC single-building owners/boards, capped, this filing season.
• Incentive: founding-client price lock + free 2030 gap estimate.
• Feedback cadence: weekly review of every filing; log each PE correction.
• Product feedback vs custom work: a recurring correction/rule = product; a one-off owner request = custom (declined or priced separately).
• Corrections → system: each fix becomes an SOP step, a rules-library entry, a prompt tweak, or a QA check.
• Gate before expanding: intake completeness >95%, rework <5%, PE minutes/unit trending down.

7 / 30 / 90-day launch plan

Metrics & KPIs

Exhaustive risk register

What could kill this

The two existential threats are (a) broad, durable political rollback of BPS penalties across most jurisdictions simultaneously — unlikely given 40+ independent laws and tightening climate mandates — and (b) failure to hold the PE-certification bottleneck economics, which would cap scaling. Both are mitigated by multi-jurisdiction diversification and by engineering the workflow so licensed time is spent only at the signature. A third, softer risk is over-indexing on penalties in the pitch; the durable value is the mandatory annual certified filing, which exists regardless of penalty levels.

Go/no-go reasoning & final recommendation

The candidate clears every element of the evidence threshold: a clearly identified buyer (covered-building owners/boards), a specific and quantified painful problem (escalating per-ton fines + mandatory certified filing), verified active spend (dozens of competing paid services, published per-building fees), competitor/budget validation, a licensing moat, a narrow one-feature MVP, a service-first delivery path, a credible route to 55%→78%+ gross margin, and multiple believable distribution channels. Outcome/per-unit pricing is natural and legal when structured carefully. No fatal disqualifier is triggered.

Recommendation: BUILD. Launch the NYC LL97 Emissions Report + Certification wedge this filing season with a partner PE, ship the free Penalty Exposure Estimate lead magnet, cap the first pilot cohort at 8–10 buildings, and add Boston/DC rules packs by day 90.

Sources

1. [S1] NYC Dept. of Buildings — LL97 Greenhouse Gas Emissions Reduction — nyc.gov
2. [S2] NYC Accelerator — Local Law 97 — accelerator.nyc/ll97
3. [S3] REBNY — Report: LL97 fines could exceed $900M/yr by 2030 — rebny.com
4. [S4] Urban Green Council — Local Law 97 — urbangreencouncil.org
5. [S5] City of Boston — Building Emissions Reduction & Disclosure (BERDO) — boston.gov
6. [S6] Institute for Market Transformation — U.S. Building Performance Standards map — imt.org
7. [S7] Facilities Dive — 2026 map of BPS across the US — facilitiesdive.com
8. [S8] NYC Dept. of Buildings — LL84 Benchmarking — nyc.gov
9. [S9] Greenberg Traurig E2 Law Blog — LL97 emissions limits take effect — gtlaw-environmentalandenergy.com
10. [S10] Colorado Energy Office — Building Performance Colorado — energyoffice.colorado.gov
11. [S11] Envigilance — Building Performance Standards: 50+ US City Laws — envigilance.com
12. [S12] 360iResearch — Decarbonization Software Market — 360iresearch.com
13. [S13] Elevate — Chicago Energy Benchmarking Compliance Services — elevatenp.org
14. [S14] EZ Energy Services — NJ Benchmarking — ezenergyservices.com
15. [S15] NYC Dept. of Buildings — LL97 Covered Buildings List (CY2025, PDF) — nyc.gov
16. [S16] Energo — NYC Local Law 97 (services + $0.50/sq ft/mo non-filing) — energo.com

Generated 2026-07-05. Claims labeled Verified / Inferred / Unverified. Market-sizing ranges are explicitly Inferred and should be treated as directional, not precise.