Overview
VacStak, powered by NumeriXGPT LLC, translates USDA-APHIS, NAHMS, and peer-reviewed veterinary vaccine research into AI-powered economic calculators for livestock producers, aquaculture operations, veterinarians, and agricultural professionals. As VacStak scales toward partnerships with extension services, veterinary schools, and federal animal health agencies, formal data governance is essential to maintaining research integrity, user trust, and institutional credibility.
Our governance framework covers three interconnected domains:
Veterinary Research Data
How USDA-APHIS, NAHMS, and peer-reviewed vaccine efficacy datasets are sourced, versioned, validated, and maintained across all 16 VS-series models.
User & Platform Data
How producer inputs, calculation outputs, herd/flock parameters, analytics events, and account information are classified, stored, and protected.
AI & Cloud Data
How AI model interactions, cloud infrastructure, and third-party veterinary APIs are governed to protect partner and user data.
VacStak's methodology for translating USDA-APHIS veterinary research into AI-powered vaccine economics calculators is patent pending under US Patent Application 63/970,943 (NumeriXGPT LLC). Data governance records — including model sourcing logs and versioning history — serve as supporting documentation for IP protection.
VacStak operates under the same governance infrastructure as AgrStak. Policies governing user PII, cloud security, AI outputs, data retention, and incident response are shared across both platforms. VacStak-specific governance covers VS-series model sourcing, veterinary data classification, and USDA-APHIS/NAHMS citation standards.
Governance Structure
As a founder-led platform, NumeriXGPT LLC consolidates all governance roles under a single accountable party during the current phase of VacStak's development. Roles will be formally separated as the team grows toward enterprise scale.
| Role | Responsibility | Current Holder |
|---|---|---|
| Data Governor | Policy authority and final decisions on all data use across VacStak and AgrStak domains | CEO / Founder |
| Veterinary Research Data Steward | USDA-APHIS and NAHMS dataset sourcing, versioning, and citation accuracy for all 16 VS models | CEO / Founder |
| Platform Data Steward | User data, privacy compliance, analytics hygiene across VacStak and AgrStak | CEO / Founder |
| AI/Cloud Data Steward | AI model inputs/outputs, cloud security, API contracts for veterinary data sources | CEO / Founder |
| External Auditor | Annual third-party compliance review | TBD — Year 2 |
Research Data Governance
VacStak's core value proposition is built on peer-reviewed veterinary vaccine economics research from USDA-APHIS, the National Animal Health Monitoring System (NAHMS), and published epidemiological literature. How that research is sourced, maintained, and cited directly affects credibility with institutional partners and the accuracy of every vaccine ROI calculator output.
Sourcing Standards
- All vaccine efficacy and economic models derived from peer-reviewed USDA-APHIS publications, NAHMS reports, or official USDA veterinary technical bulletins
- Source documents archived internally at the time of each VS model's creation
- No proprietary third-party pharmaceutical or vaccine manufacturer datasets incorporated without a written licensing agreement
- Model inputs and coefficients — including morbidity rates, mortality rates, treatment costs, and vaccine efficacy parameters — must be traceable to the original USDA or peer-reviewed source
- Epidemiological models (VS11) and biosecurity models (VS12) sourced from USDA-APHIS Veterinary Services publications and peer-reviewed biosecurity literature
- Aquaculture vaccine models (VS7–VS9) sourced from USDA-ARS and NOAA Fisheries veterinary research where applicable
Model Versioning
Each VS calculator model is tied to a specific version of the underlying veterinary research. When USDA-APHIS or NAHMS updates its findings, VacStak updates the corresponding model. Every model in the VS registry carries the following metadata:
| Field | Description | Required |
|---|---|---|
| Model ID | Unique VS-series identifier (e.g., VS1-6005, VS11-6015) | Yes |
| Source Citation | Full APA citation of USDA-APHIS, NAHMS, or peer-reviewed source publication | Yes |
| Version Number | Semantic version of the model (e.g., v1.0.0) | Yes |
| Last Validated | Date model was reviewed against current USDA-APHIS or NAHMS data | Yes |
| Research Contact | USDA-APHIS or NAHMS researcher who authored the source publication | Recommended |
| Species Scope | Livestock/aquaculture species covered by the model | Yes |
| Deprecation Date | Date model will be retired if not updated with current veterinary research | Yes |
All VS models undergo internal validation before deployment. Material discrepancies greater than 5% deviation from USDA-APHIS or NAHMS source data trigger a model review before launch. All 16 VS models are reviewed annually against the latest USDA-APHIS and NAHMS publications. Vaccine efficacy parameters are additionally reviewed whenever APHIS publishes updated disease incidence data.
VS Model Registry
VacStak's 16 VS-series models span general vaccine economics, livestock, aquaculture, honeybee, epidemiology, and biosecurity. Each model carries a globally unique numeric identifier in the 6000-series, shared across Science and Economics modules within the same storefront.
| VS Series | Model ID | Domain | Species / Scope | Status |
|---|---|---|---|---|
| VS0 | 6001–6004 | General Vaccine Economics | Multi-species — Free Combined Calculator | Deployed |
| VS1 | 6005 | Livestock | Cattle (beef & dairy) | Deployed |
| VS2 | 6006 | Livestock | Swine | Deployed |
| VS3 | 6007 | Livestock | Sheep & Goat | Deployed |
| VS4 | 6008 | Poultry | Broiler & Layer Chicken | Deployed |
| VS5 | 6009 | Poultry | Turkey | Deployed |
| VS6 | 6010 | Avian Health | Avian (expanded — wild bird interface) | Deployed |
| VS7 | 6011 | Aquaculture | Shrimp | Deployed |
| VS8 | 6012 | Aquaculture | Salmon | Deployed |
| VS9 | 6013 | Aquaculture | Tilapia | Deployed |
| VS10 | 6014 | Apiculture | Honeybee Colony Health | Deployed |
| VS11 | 6015 | Epidemiology | Herd/Flock Epidemiology Modeling | Deployed |
| VS12 | 6016 | Biosecurity | Farm Biosecurity Economics | Deployed |
The next available VS-series model identifier is 6017. All new VacStak calculator storefronts must register with a globally unique 6000-series ID prior to deployment. Science (S) and Economics (E) modules within the same storefront share the same model number.
User & Platform Data Governance
VacStak collects data from livestock producers, aquaculture operations, veterinarians, and agricultural professionals. This section governs what is collected, how it is stored, and how it is protected.
Data Classification
| Class | Examples | Protection |
|---|---|---|
| Public | VS calculator descriptions, USDA-APHIS model summaries, published vaccine efficacy citations | Standard |
| Internal | Business logic, pricing, model coefficients, internal analytics, VS model registry | Restricted |
| Confidential | Partner contracts, LOIs, unpublished USDA-APHIS research, DSA data, veterinary school agreements | Highly Restricted |
| User PII | Names, emails, operation location, herd size, subscription records | Maximum Protection |
| Calculator Inputs | Herd size, vaccination rates, mortality parameters, species-specific production data | Ephemeral — session only, never stored |
VacStak calculator inputs — including herd size, species count, vaccination rates, mortality rates, and production economics — are processed in-browser and are never transmitted to or stored on VacStak servers. All economic calculations occur client-side. No farm or operation data leaves the user's device.
Data Retention
AI & Cloud Governance
VacStak deploys AI features including inline chat assistants (ACW), AI-generated Detailed Reports, and research interpretation tools across VS storefronts. This section governs how AI systems interact with user data and veterinary research content.
Research-Grounded Outputs
AI outputs in VacStak are always grounded in the USDA-APHIS and NAHMS research underlying each VS model. AI systems do not generate novel veterinary recommendations outside the scope of the source research.
No Input Data Training
User calculator inputs — herd size, vaccine parameters, production economics — are never used to train or fine-tune AI models. VacStak does not share user operational data with any AI provider for training purposes.
Transparent Uncertainty
AI-generated Detailed Reports and interpretive summaries clearly distinguish between modeled outputs (based on USDA research) and AI-generated narrative. Uncertainty ranges are disclosed where present in source data.
Not a Veterinary Diagnosis
VacStak calculators and AI outputs are economic and research-support tools only. They are not a substitute for licensed veterinary advice, USDA-APHIS regulatory guidance, or a formal herd health plan. All outputs carry this disclaimer.
Third-Party AI Providers
VacStak's AI features are powered by Anthropic's Claude API. Our data processing agreement with Anthropic governs how prompts and outputs are handled. Key provisions:
- Anthropic does not use API inputs or outputs to train Claude models by default
- Calculator inputs are never included in AI prompts — only model metadata and research context are passed to the AI layer
- All AI API calls use HTTPS with TLS 1.3 encryption in transit
- API keys are stored as environment variables and never exposed in client-side code
Cloud Infrastructure
- VacStak is deployed on Vercel with GitHub-based CI/CD pipelines
- All secrets and API keys managed via environment variable injection — never hardcoded
- GitHub repository is private; access restricted to authorized contributors
- Automatic dependency vulnerability scanning enabled via GitHub Dependabot
- Infrastructure shared with AgrStak; security controls apply across both platforms
Incident Response
In the event of a data security incident — including unauthorized access, data exposure, or system compromise — VacStak follows a structured six-step response protocol shared with AgrStak under the NumeriXGPT LLC Incident Response Policy.
| # | Step | Action | Timeline |
|---|---|---|---|
| 1 | Detect | Identify and confirm the incident through monitoring, user report, or internal discovery | Immediate |
| 2 | Contain | Isolate affected systems, revoke compromised credentials, preserve evidence | Within 4 hours |
| 3 | Assess | Determine scope, classify data types exposed, evaluate impact on users and institutional partners | Within 24 hours |
| 4 | Notify | Notify affected users, veterinary institutional partners, and regulators as required by law | Within 72 hours |
| 5 | Remediate | Fix root cause, implement additional controls, restore VacStak services | Within 7 days |
| 6 | Review | Post-incident report documenting cause, impact, and prevention measures | Within 30 days |
In the event of a data incident involving Shared Data governed by a Data Sharing Agreement (DSA) with a veterinary school, extension service, or USDA-APHIS institutional partner, the affected party will be notified within 72 hours of discovery per the DSA breach notification clause.
Governance Review Calendar
Data governance is not a one-time activity. VacStak maintains a structured review cadence to keep the framework current and effective — synchronized with the AgrStak governance calendar under NumeriXGPT LLC.
Analytics Review
GA4 data quality, event tracking accuracy, anomaly detection across all VS storefronts.
Security Audit
API key rotation, GitHub secret scanning, vendor access review, and cloud configuration check — shared with AgrStak.
AI Output Review
Flagged AI outputs reviewed, hallucination rate assessed, user feedback patterns analyzed across VS chat interfaces.
Full Framework Review
All policies updated, USDA-APHIS and NAHMS model validation completed, privacy policy refreshed.
Vendor Review
All third-party Data Processing Agreements reviewed. New veterinary data vendors assessed before integration.
DSA Execution
A custom Data Sharing Agreement executed with each new veterinary or institutional partner prior to data exchange.
Governance Roadmap
As VacStak scales alongside AgrStak, governance infrastructure grows with it. The following milestones define our path from startup to enterprise-grade veterinary data governance.
Foundation
Publish this VacStak Data Governance Framework publicly. Complete the internal VS model registry for all 16 calculators (VS0–VS12). Execute first DSA with a USDA-APHIS or veterinary school institutional partner.
Section 508 Compliance Audit
Complete WCAG 2.1 AA accessibility audit across all VS storefronts. Remediate any barriers to ensure the platform serves the full spectrum of livestock producers and veterinary professionals.
Team & Role Separation
Hire a dedicated Platform Data Steward as the team grows. Separate veterinary research governance from user data and AI/cloud governance roles.
SOC 2 Type I Certification
Pursue SOC 2 Type I (shared with AgrStak) to meet the security requirements of veterinary school, extension service, and federal agency institutional customers.
SOC 2 Type II & External Audit
Achieve SOC 2 Type II certification. Establish a formal annual external audit cycle with a qualified third-party auditor covering both VacStak and AgrStak.
Automated Data Lineage
Implement automated data lineage tracking across all 16 VS storefronts — tracing every vaccine efficacy parameter from USDA-APHIS or NAHMS source to calculator output.