WVPPClear — AI-Native California SB 553 Workplace Violence Prevention Plan Engine

01 Thesis

Since July 1, 2024, most California employers must establish, implement, and maintain a site-specific written Workplace Violence Prevention Plan, keep a violent incident log, train employees, and retain hazard/training/investigation records under Labor Code §6401.9 (SB 553). Cal/OSHA already enforces the statute and must adopt a permanent general-industry standard by December 31, 2026. Cheap template SaaS exists, but Cal/OSHA expects plans specific to each work area and operation — a documentation and hazard-synthesis burden that mid-market multi-site HR/safety teams still scramble to fulfill. WVPPClear sells a specialist-QA’d, site-specific WVPP compliance package as a done-for-you service: AI drafts and maps; humans own hazard completeness and site-specificity QA; the customer remains the employer of record. AI is the internal production engine, not the product the buyer operates.

02 Discovery rationale

Terrain scanned this hour: California SB 553 / LC 6401.9 WVPP, FERC Order 917 EQR XBRL transition, EPA Subpart W / GHGRP (rescission risk), TSCA §12(b) export notification, Form LM-10 employer reporting, industrial MSGP annual reports, EPA RMP SCCAP (rollback proposed), DEA ARCOS, OSHA ITA (already claimed), EPCRA Tier II (already claimed inside prior EPCRA/TRI engine), and CARB Advanced Clean Fleets (high-priority components repealed/in flux). Prior OFF-LIMITS (EPRCClear, DROPClear, MCASClear, PipeClear, Cert500, CleanBuildClear, BERDOClear, intl-tax stack, elective pay, NYDFS Part 500, PHMSA, CA pay-data, privacy DSAR/DROP adjacency, insurance retaliatory/MCAS) were excluded. WVPPClear won because it combines (1) a live statewide mandate already enforceable since July 2024, (2) a hard 2026 permanent-standard remapping deadline that creates a second purchase wave, (3) verified public competitor pricing and attorney spend proving category budget, (4) a service-first wedge that does not require building a safety SaaS platform before revenue, and (5) material differentiation from prior osha-recordkeeping-ita-reporting-engine (Cal/OSHA WVPP vs federal OSHA electronic Form 300A/300/301 ITA submission).

03 Candidate comparison

Scoring uses the 15-dimension 1–5 rubric (max 75). Full gate detail for the winner is in §05. Subpart W GHGRP and RMP SCCAP were screened and dropped early due to active EPA rescission/rollback proposals.

04 Hard disqualifier check

05 Rubric scorecard

Average gates: 4.17 / 5. Buyers already purchase Cynserus/Kestralis/attorney WVPP help and care about a Cal/OSHA-ready package, not tooling. Most steps are structured drafting/mapping; true judgment is site-hazard completeness and whether the plan is specific to each work area. Regulation creates recurring annual reviews and a 2026 permanent-standard remapping wave. Physical labor is optional (remote MVP). Models improve drafting throughput; the company still wins via multi-site portfolio ops, inspection-response packets, and standard-change remaps.

06 Opportunity

CODE validation

07 Evidence quality and source-claim matrix

08 Why now

• Verified Mandate has been enforceable since July 1, 2024 — many employers still have generic or incomplete programs.
• Verified Permanent Cal/OSHA standard must be adopted by Dec 31, 2026, forcing plan remaps and creating a second purchase wave.
• Verified 2025 penalty maxima rose (general/regulatory $16,285; willful/repeat $162,851), raising the cost of documentation gaps.
• Verified Public competitor category proves budget exists from $249 SaaS to $5k concierge / $3–5k attorney work.
• Inferred Frontier models now make site-questionnaire → plan drafting assistive enough that a small specialist team can productize what used to be attorney/custom consulting hours.
• Unverified Hypothesis that multi-site mid-market will pay DFY retainers rather than operate SaaS — test in pilots.

09 Customer & PMF

First one-feature MVP wedge

• ICP: CA employer with 3–8 similar worksites (e.g., retail/restaurant/warehouse) and one HR owner
• Trigger event: Annual review due, counsel request, or inspection anxiety within 60 days
• Pain: Corporate template is not site-specific; incident log/training records incomplete
• One-feature MVP: One site-specific WVPP compliance package per site (batch of 3–5 sites)
• Input: Site roster, NAICS/operations description, floorplan/photos, prior IIPP/WVPP if any, manager hazard questionnaire, reporting channel details
• Output: Site-specific WVPP PDF, hazard ID/evaluation/correction matrix, violent incident log workbook + field schema, training outline + attendance sheet, annual review calendar, Cal/OSHA inspection response checklist, evidence binder
• Human chokepoint: Safety/compliance specialist QA on site-specificity and hazard completeness; customer signs acceptance and implements
• Success metric: Customer accepts package; specialist minutes ≤ budget; zero critical LC 6401.9 element gaps on QA checklist
• Next asks if wedge works: Multi-site retainer; training delivery; incident-log portal ops; 2026 permanent-standard remap; inspection response retainer

10 The outcome we sell

• Deliverable: Site WVPP, hazard matrix, log workbook, training kit, review calendar, inspection checklist, evidence binder
• Acceptance criteria: Specialist QA sign-off against LC 6401.9 element checklist; customer confirms site facts accurate
• Customer promise: Package covers required written-plan elements and supporting kits for the named site; not a guarantee of zero citations or prevention of violence
• Exclusions: Legal representation; on-site security staffing; healthcare §3342 hospital WVPP specialty (separate product); federal OSHA ITA filing
• Refund/rework: One free revision cycle within 14 days for factual site corrections; no refund after acceptance
• Success metric: Accepted package + element-completeness score ≥98% on internal checklist

11 Internal AI engine architecture

1. Intake layer: Secure upload of site roster, questionnaires, photos/floorplans, prior IIPP/WVPP, reporting-channel details, training history.
2. Normalization layer: Structure sites, roles, work areas, visitor/public exposure flags, prior incidents, existing controls; version all inputs.
3. Retrieval/knowledge layer: Index LC 6401.9 elements, DIR model plan sections, industry hazard libraries (retail robbery, late-night, cash handling, patient/public aggression, warehouse vehicle/pedestrian, etc.), gold-standard packages, prior QA corrections.
4. AI workbench: Draft WVPP sections; map questionnaire answers to required elements; propose hazard ID/evaluation/correction rows; generate training outline; build log field schema; flag missing evidence.
5. Deterministic rules: Required-element completeness gates; retention period checks; site-specificity checks (work-area names present; no “corporate only” language); annual-review date math; prohibited-claim language filters.
6. Human chokepoint: Safety/compliance specialist reviews hazard completeness, site-specificity, and high-risk edge cases; escalates legal questions to counsel (customer’s or fractional).
7. QA layer: Dual checklist (statute elements + red-team “generic plan” detector); confidence scoring; audit trail of edits.
8. Delivery layer: PDF/DOCX package + spreadsheet kits + customer acceptance form + implementation runbook.
9. Learning loop: Capture specialist edits, inspection feedback, customer corrections into templates/rules/retrieval.
10. Model-portability: Prompt/tooling abstracted so drafting/classification models can be swapped as frontier models improve.

12 AI-vs-human operations pipeline

What the human expert actually does

13 Operations as product

• SOP library per industry vertical (retail, restaurant, warehouse, clinic, property mgmt).
• Structured intake checklist + required evidence list (photos of entrances, cash areas, panic buttons, lighting, visitor controls).
• Automated completeness checks before specialist queue.
• Exception queues: missing site facts, high-risk public-facing ops, union/work-rule conflicts, healthcare carve-outs.
• Reviewer assignment by vertical; confidence scoring on AI drafts.
• Version control of every WVPP; gold-standard examples per vertical.
• Red-team “generic plan” detector; postmortem on any inspection finding or customer rejection.

14 No-holes quality engine

• Statute-element checklist mapped to LC 6401.9 required procedures.
• Site-specificity gate: named work areas, local corrective measures, local reporting path.
• Log schema gate: fields for date/time/location/type/consequences/response.
• Training kit gate: initial + annual outline; attendance sheet; retention note (≥1 year).
• Hazard record retention note (≥5 years) embedded in binder.
• Prohibited claims filter: no “Cal/OSHA certified,” no guarantee of zero violence/citations, no legal advice.
• Human sign-off required before delivery; customer acceptance required before close.

15 Pricing, pricing legality, and unit economics

• Primary pricing: Per-site WVPP package Inferred $1,800–$4,500; multi-site portfolio (5 sites) $8,000–$15,000; annual review/remap retainer $150–$400/site/year after year 1.
• Why not hourly: Buyers want a finished package and inspection-ready binder, not billable hours.
• Pricing legality: Fixed fees for documentation preparation are appropriate. No contingency on avoided fines. Do not sell “citation insurance.”
• COGS per site @ launch (inferred): model/doc processing $3–8; software/hosting $5–15; specialist ~90–130 min (~$75–$150 labor); QA/ops $15–30; total COGS ~$100–$200 → gross margin target 70%+ at $2,500 ASP.
• Automation: ~55% launch → ~75% day 90 → ~85% year 1 on drafting/completeness; specialist remains on hazard QA.
• Revenue/FTE target: $350k–$500k after process hardening.

16 Nonlinear scaling plan

• Throughput target: 3–4 sites/specialist/day at day 90 after libraries mature.
• Cycle time: 5–7 business days launch → 2–3 days day 90 for standard verticals.
• Rework <8%; escalation <10%; quality failure (critical element miss) <1%.
• Margin expands as vertical templates, photo-check rules, and generic-language detectors absorb specialist minutes.
• Do not scale by adding custom on-site consulting for every site; productize remote evidence standards first.

17 Moat & Sam Altman test

Frontier models make drafting cheaper — that helps WVPPClear’s COGS more than it helps a one-off DIY employer, because the company owns vertical hazard libraries, multi-site calendars, inspection-packet playbooks, and the 2026 permanent-standard remap workflow. Strongest commoditization threat: $29–$249/yr template SaaS. Anti-commoditization: Cal/OSHA site-specificity requirement + multi-site portfolio ops + specialist QA + annual review/remap + 48-hour inspection response retainer. The product is the operating system for ongoing compliance, not a PDF generator login.

18 Buyer-specific go-to-market

Selected motion: Founder-led outbound to CA multi-site mid-market + free WVPP gap diagnostic + employment counsel / PEOs / insurance broker referrals. Waitlist/lead magnet fit because buyers search “SB 553 compliance” and will exchange site facts for a gap score.

Distribution proof table

First 30 days of content

• 10 posts: site-specificity vs corporate templates; violent incident log mistakes; training record retention; complaint-triggered inspections; 2026 permanent standard; retail late-night hazards; warehouse aggression; restaurant cash-handling; multi-site calendar failures; what Cal/OSHA asks for first.
• 3 teardowns: anonymized generic-plan fails; missing log fields; incomplete annual review.
• 2 lead magnets: 22-point WVPP gap checklist; multi-site readiness scorecard.
• 1 webinar: “Make every CA site inspection-ready before the 2026 standard.”
• 1 outbound diagnosis template: 8-question site gap memo with severity tags.

Lead magnet & waitlist

Free WVPP Gap Diagnostic: scored checklist + site-specificity findings. Waitlist CTA for multi-site portfolio pilots (cap 5). Follow-up within 1 business day; sales-ready if ≥3 sites and ≥2 critical gaps.

19 Pilot design and early-demand trap mitigation

• Pilot cap: 5 employers, ≤5 sites each (max 25 site-packages).
• Incentive: 30% off year-1 package in exchange for weekly feedback + anonymized teardown rights.
• Success criteria: ≥80% acceptance without major rework; specialist ≤120 min/site by site #15; element-completeness ≥98%.
• Product feedback vs custom work: Feedback = checklist misses, template gaps, intake friction. Custom = unique security engineering, legal opinions, on-site guarding — refuse or refer.
• Build-before-scale: After 5 customers harden intake/evidence; after 10 harden SOPs/exception queues; after 20 pause until COGS, rework, cycle time measured.

20 Competitive landscape

• Low-touch SaaS: Cynserus, CompliantCA, SB553Ready — fast PDFs/portals; buyer still operates compliance.
• Advisory/DFY: Kestralis Complete/Concierge; local EHS consultants; employment counsel ($3–5k+).
• Enterprise EHS suites: NAVEX/Resolver-class — overkill/price for mid-market SB 553-only need.
• DIY: Cal/OSHA model plan — free but generic risk.
• Why WVPPClear wins: DFY multi-site site-specificity + specialist QA + annual remap ops without enterprise suite cost; not another self-serve template login.

Anti-duplication analysis

• Not a clone of federal OSHA ITA filing (different regulator, forms, outcome).
• Not generic AI consulting or “AI for HR.”
• Not a directory/content site.
• Narrow wedge: CA multi-site LC 6401.9 packages with specialist-reviewed site hazard synthesis.
• Unsolved by SaaS: portfolio ops, inspection packets, 2026 remaps, enforcing site-specificity at scale.

21 Regulation, compliance, and licensing boundary

• AI may: draft WVPP text, extract site facts, classify hazards from intake, generate checklists/logs/training outlines, flag missing elements.
• Operators may: QA completeness/site-specificity, package deliverables, run calendars, coach implementation logistics.
• Licensed professionals: Customer’s employment counsel for legal opinions; CSP/ASP optional for complex hazard programs; do not practice law or claim Cal/OSHA certification authority.
• Must not claim: legal advice; guaranteed citation prevention; violence prevention warranty; healthcare hospital WVPP under separate standards without specialty process.
• Required controls: engagement letter, employer-responsibility disclaimer, audit logs, no contingency-on-fines pricing.

22 Compact founding team and expert map

23 Exhaustive risk register

24 Tech stack & build plan

1. Week 1–2: Typeform/Airtable intake, Drive templates, checklist spreadsheet, Doc AI drafting prompts, e-sign acceptance.
2. Week 3–6: Internal workbench (Next.js + Postgres): site objects, element gates, specialist queue, versioning.
3. Day 90: Vertical libraries, generic-language detector, calendar automations, partner portal.
4. Later: optional customer incident-log portal (support layer only).

25 Metrics & KPIs

• Throughput sites/specialist/day; cycle time; rework %; escalation %; automation %; element-completeness %; customer acceptance %; COGS/site; gross margin; revenue/FTE; diagnostic→pilot %; pilot→paid %; annual retainer renewal %.

26 What could kill this

• Inability to sell above SaaS price without on-site labor.
• Permanent standard so template-friendly that site-specificity moat collapses.
• Liability event from a missed hazard without enforceable disclaimers/E&O.
• Pilot custom-work trap destroying unit economics.

27 90-day validation and launch plan

• Days 1–7: Finalize LC 6401.9 checklist, 2 vertical templates, engagement letter, diagnostic landing page; outreach 50 prospects.
• Days 8–30: Close ≤5 pilot employers; deliver first 10 sites; instrument minutes/rework; publish 10 content pieces.
• Days 31–60: Harden intake/QA; add 2 verticals; test $2,500/site and 5-site $10k offers; sign 2 counsel/broker partners.
• Days 61–90: Measure COGS/margin; pause new pilots if rework >15% or specialist >150 min/site; decide go/expand/kill.
• Kill criteria: <3 paid pilots; >30% require unpaid on-site; cannot clear 50% gross margin path; UPL/liability issues unresolved.

7-day / 30-day launch snapshots

7-day: checklist + retail/restaurant templates + diagnostic + 50 outbound. 30-day: 5 pilots contracted, ≥8 sites delivered, first teardown published, pricing validated in writing.

28 Sources

1. California LegInfo — SB 553 bill text
2. DIR Cal/OSHA — Workplace Violence Prevention General Industry
3. DIR — Laws in General Industry (LC 6401.9 summary)
4. DIR — Workplace Violence Prevention guidance hub
5. DIR — Employer fact sheet (WVPP)
6. DIR News — Cal/OSHA civil penalty amounts 2025
7. Jackson Lewis — Cal/OSHA SB 553 FAQs summary
8. Cynserus pricing
9. Cynserus — SB 553 enforcement overview
10. Kestralis SMB SB-553 plans
11. Kestralis — What Cal/OSHA looks for in WVPP inspections
12. CompliantCA
13. SB553Ready
14. PPIC — California’s Businesses
15. FERC Order 917 (rejected adjacent candidate)
16. EPA GHGRP rulemaking notices
17. EPA RMP Common Sense proposed rule
18. EPA TSCA export requirements
19. DOL OLMS employer/consultant reporting (LM-10)
20. EPA 2021 MSGP permit (annual report context)

Run 20260711-00-01 · Hard-to-fool blueprint · Claims labeled Verified / Inferred / Unverified · PDF generated from this HTML.