Security & Architecture

Enterprise security,
embedded by design.

StandardPulse Security & Architecture Whitepaper

Last Updated: March 2026

Executive Summary

StandardPulse is a modern GRC intelligence platform built for monitoring, triaging, and managing compliance activities across global jurisdictions.

Because our platform is utilized by vCISOs, compliance architects, and enterprise risk teams, security and data privacy are not layered on top of our application—they are embedded into the foundational architecture.

This document outlines the system architecture, data isolation strategies, and artificial intelligence safeguards implemented within StandardPulse to ensure enterprise-grade security and confidentiality.

AI Safeguards

Artificial Intelligence & Data Privacy

One of the most critical concerns for modern GRC teams is the handling of proprietary control frameworks and sensitive gap analyses by Large Language Models (LLMs). StandardPulse employs a strict, enterprise-first approach to AI.

Zero-Retention Enterprise APIs

All AI processing is powered by the Google Gemini API (@google/genai). StandardPulse utilizes enterprise API endpoints, meaning your proprietary data, custom frameworks, and internal SME comments are never used to train foundational AI models.

Prompt Injection Hardening

All user-supplied content processed by the AI engine is sandboxed within strict delimiter tags with explicit system-level security instructions that mitigate adversarial prompt injection or role override. AI outputs are further sanitized server-side through field-length caps, enum whitelisting, and schema validation before being persisted, ensuring that neither user input nor model output can corrupt application state.

Hybrid Deterministic + AI Matching

To virtually eliminate AI hallucinations in compliance mapping, our Compass and STRM (Set Theory Relationship Mapping) tools do not rely solely on generative text. We utilize a multi-tier engine, with human-in-the-loop operations:
  • Tier 1 (Deterministic): Exact citation and fuzzy matching based on the Secure Controls Framework (SCF).
  • Tier 2 (Semantic): Vector embedding similarity to retrieve specific, verifiable controls before the LLM generates a classification and confidence score.
  • Tier 3 (Human-in-the-Loop): Top-3 vector embedding results are presented to the human expert to assess and determine the best control to map.

NIST-Backed Methodologies

Our automated requirement decomposition and classification pipeline is strictly modeled after NIST IR 8477 (Set Theory Relationship Mapping), ensuring audit-defensible rationale (chain-of-thought) for every mapping.

System Flow

Platform Architecture

A high-level look at how StandardPulse securely handles data from the client, through our isolated middleware, and into our hardened storage and AI layers.

Client Workspaces

Encrypted TLS 1.3 Connection

API Gateway & Identity Middleware

Google SSO Auth • Tenant Boundary Validation (Org ID)
Express Rate Limiting • Zod Input Sanitization

Isolated Data Store

AES-256 Encryption
Strict Row-Level Security
Continuous Snapshots

AI Processing Pipeline

Sandboxed Delimiters
Deterministic Vector Matching
Zero-Retention Enterprise API

Architecture

Multi-Tenant Organization Isolation

StandardPulse utilizes a robust multi-tenant architecture to ensure absolute segregation of client and organizational data.

Cryptographic Tenant Isolation

Every piece of sensitive data—including tracked sources, STRM runs, radar configurations, custom frameworks, and intake queues—is strictly bound to a unique organization_id.

API Middleware Enforcement

Our API routes are protected by tenant-aware middleware. Every request validates the authenticated user's organization_id and role before database execution, preventing cross-tenant data bleed.

Organization Lifecycle Management

Tenant environments are tightly controlled through explicit lifecycle states (active, read_only, inactive, dormant, deleted) to ensure data is properly retained or purged according to customer directives and industry standard-based account lifecycle management controls.

Access Control

Identity & Access Management (IAM)

Authentication and authorization are managed through industry-leading identity providers and granular Role-Based Access Control (RBAC).

Bank-Grade Authentication

Identity management is offloaded to Google Cloud Identity Platform, supporting robust Google SSO and secure token-based authentication (Authorization: Bearer <token>).

Granular RBAC

Access within an organization is strictly governed by distinct roles:

  • Owner / Admin: Full platform control, billing management, and source list modification.
  • Member: Standard operational access (triaging, mapping, commenting).
  • Viewer: Read-only access to dashboards, reports, and mappings.

Secure Invites

Organization expansion is handled via time-limited, cryptographically secure invite tokens.

Infrastructure

Application Security & Resilience

Our application stack is hardened against common web vulnerabilities and operational disruptions.

Runtime Protection

The API is fortified using Helmet for secure HTTP headers, strictly configured CORS policies, and rigorous input sanitization via Zod schema validation to prevent injection attacks. AI-generated outputs are additionally sanitized through server-side field-length caps, tag whitelisting against allowed values, and numeric range clamping before persistence.

Rate Limiting & Abuse Prevention

All API endpoints—particularly heavy AI-generation and document-extraction routes—are protected by express-rate-limit with automatic request throttling to ensure high availability and mitigate Denial of Service (DoS).

Data Resilience & Integrity

StandardPulse enforces strict database-level integrity and automated snapshotting to ensure high availability, concurrent read/write safety, and rapid recovery capabilities.

Continuous Vulnerability & Posture Management

To proactively identify and mitigate risks, StandardPulse integrates native Google Cloud security guardrails directly into our deployment lifecycle. All application container images and dependencies undergo automated vulnerability scanning (CVE detection) via Google Artifact Registry. Furthermore, our production environment is continuously monitored by Google Security Command Center to detect infrastructure misconfigurations and enforce strict cloud security posture management, ensuring potential flaws are identified and patched before they can be exploited.

Audit Logging & Observability

StandardPulse maintains a compliance-grade audit trail that automatically records all security-relevant operations—including authentication events, data mutations, administrative actions, and AI processing requests—with structured JSON output. Every log entry captures best-practice audit metadata: who performed the action, what was affected, when it occurred, where the request originated, and why it was classified at a given severity. All structured logs are auto-ingested by Google Cloud Logging with severity mapping (INFO, WARNING, ERROR, CRITICAL), enabling centralized querying, automated alerting, and SIEM integration. Audit records are retained for 180 days with automated purging to optimize storage costs while maintaining compliance with SOC 2 and NIST AU-2/AU-3 requirements.

Formal Compliance Policies

StandardPulse maintains formal, framework-aligned compliance policies covering Access Review (quarterly reviews, account lifecycle, revocation SLAs), Incident Response (5-phase IR with severity classification and communication procedures), Business Continuity & Disaster Recovery (RPO ≤1 hour, RTO ≤2 hours, documented recovery procedures), and Vulnerability Management (CVSS-based remediation SLAs, CI/CD scanning gates, risk acceptance). All policies are mapped to SOC 2, ISO 27001, and NIST 800-53 controls.

Billing & Infrastructure Security

Isolated Payment Processing

StandardPulse does not store or process raw credit card data. All subscription, checkout, and billing operations are securely offloaded to Paddle via secure webhook integrations with HMAC-SHA256 signature verification.

Cloud Infrastructure

The platform is deployed entirely on Google Cloud Platform (GCP). Data is encrypted in transit using TLS 1.3 and at rest using AES-256 encryption.

Data Ownership & Portability

StandardPulse believes in earning your business through ongoing value, not vendor lock-in.

Unrestricted Exports

Customers retain full ownership of their compliance data. Every STRM crosswalk, control gap analysis, AI-generated executive briefing, and SME comment review can be exported instantly to standard, audit-ready formats (XLSX, CSV, or PDF).