Track 02: Integrate

API Infrastructure.

Programmatic access to the full intelligence pipeline. Your existing systems call the API. The pipeline returns evidence-locked, scenario-complete, audit-sealed outputs. Your architecture stays intact.

REST + gRPC mTLS Async + Webhooks
Two Tracks

Build or integrate. Same pipeline.

Whether you need a complete application or programmatic access to the intelligence layer, the underlying pipeline is identical. The difference is the interface.

Track 1: Build

Custom Intelligence Applications

We build the application. Your team operates it. The intelligence layer is embedded from day one. Every output is evidence-locked, scenario-complete, and audit-sealed.

Architecture Layers
4 layers
01
Your Application

Custom interface designed for your domain

Active
02
Intelligence Core

Five-layer KRYOS reasoning pipeline

Active
03
Domain Models

Calibrated to your regulatory context

Active
04
Governance Layer

Audit trails and compliance serialization

Active
Pipeline active · All layers operational
application.ts
// Your custom application
const analysis = await kryos.analyze({
  domain: "regulatory-compliance",
  evidence: documentSet,
  scenarios: ["base", "adverse"],
  confidence_threshold: 0.92,
});

// Every output includes audit trail
console.log(analysis.audit_hash);
// → "sha256:7f3a..."
Deliverables
Domain-specific reasoning models
Full-stack application development
Cryptographic verification layer
Human escalation protocols
Ongoing calibration and support
Primary Audience
Managing PartnersGeneral CounselSovereign Entities

Same Intelligence Core

Both tracks use the same five-layer pipeline. The difference is who writes the application code. The reasoning quality, governance standards, and audit guarantees are identical.

Endpoints

Core API reference.

Five endpoints cover the complete pipeline: submit evidence, retrieve audit trails, verify independently, inspect scenario branches, and trigger human escalation.

POST/v1/analyze

Submit evidence for full pipeline analysis. Returns scenario branches, conflict map, and synthesized output with confidence score.

Parameters
evidence_set: Evidence[]
scenario_config?: ScenarioConfig
confidence_threshold?: number
Returns
AnalysisResult with branches, conflicts, synthesis, audit_hash
Response Format

Every response includes a confidence score.

The confidence score is a composite metric derived from evidence quality, scenario agreement, and conflict resolution completeness. It is not a probability. It is a measure of how well-supported the output is by the evidence.

78System ConfidenceAbove Threshold
30s agoThreshold: 65%Now
Security

Infrastructure-grade security.

Security is not a feature. It is a structural requirement. Every layer of the API infrastructure enforces authentication, encryption, and audit logging by default.

mTLS Authentication

Mutual TLS with certificate pinning. Every API call is authenticated at the transport layer before reaching the application.

Zero-Trust Network

No implicit trust between services. Every request is authenticated, authorized, and encrypted regardless of network position.

Encrypted at Rest

All data is encrypted at rest using AES-256. Encryption keys are managed through a dedicated key management service.

Rate-Limited & Throttled

Per-client rate limiting with configurable thresholds. Burst protection prevents resource exhaustion under load.

Audit-Sealed Responses

Every API response includes the audit hash for the analysis. Clients can verify the response was not modified in transit.

Webhook Callbacks

Asynchronous analysis results delivered via signed webhooks. Callback payloads include the audit hash for verification.

Quick Start

Three lines to your first analysis.

analysis.py
from kryos import Client

client = Client(api_key="your_key", endpoint="https://api.kryos.io")

# Submit evidence and receive full pipeline analysis
result = client.analyze(
    evidence=[
        {"source": "financial_report_q3.pdf", "type": "document"},
        {"source": "market_data_feed", "type": "live_feed"},
        {"source": "regulatory_filings", "type": "database"},
    ],
    config={
        "scenario_branches": ["base", "upside", "downside", "adversarial"],
        "confidence_threshold": 0.75,
        "escalation_enabled": True,
    }
)

# Access the synthesized output
print(result.synthesis.conclusion)       # Evidence-locked output
print(result.synthesis.confidence_score)  # Composite confidence
print(result.audit_hash)                  # Verify independently

# Inspect individual scenario branches
for branch in result.branches:
    print(f"{branch.type}: {branch.conclusion}")
    print(f"  Evidence: {len(branch.citations)} sources cited")
    print(f"  Confidence: {branch.confidence}")

# Retrieve the full audit trail
audit = client.get_audit(result.analysis_id)
print(f"Audit records: {len(audit.records)}")
print(f"Chain valid: {audit.verify()}")

Request
API access.

API access is provisioned after an architecture review. We will assess your integration requirements, configure the pipeline for your domain, and provision credentials.