AI Processing Disclosure

How Kyklos360 uses AI — and what it does with your data

This page is intended for enterprise procurement, legal, and compliance teams evaluating Kyklos360. Questions: info@tekcapitol.com

Section 1 — How Kyklos360 Uses AI

Kyklos360 uses large language models (LLMs) to analyze the workflow information you provide and generate your readiness report.

Specifically, AI is used to:

· Decompose your agent workflow into discrete steps
· Diagnose each step across 7 readiness dimensions
· Generate a phased remediation roadmap
· Recommend model architecture per workflow step
· Produce NIST AI RMF governance mapping
· Generate kill switch and monitoring specifications

AI analyzes your workflow data. It does not store, index, or share it beyond generating your report.

Section 2 — AI Providers

TekCapitol uses large language models from the following providers:

Google LLC
Product family: Gemini models
Role: Primary analysis engine
Subprocessor agreement: In place
Privacy: policies.google.com/privacy
Anthropic, PBC
Product family: Claude models
Role: Alternative analysis
Subprocessor agreement: In place
Privacy: anthropic.com/privacy
OpenAI, Inc.
Product family: GPT models
Role: Alternative analysis
Subprocessor agreement: In place
Privacy: openai.com/policies/privacy-policy

The specific model version used may change without notice as we optimize for quality, speed, and cost. The providers listed above will not change without updating this disclosure.

We do not use models that retain customer data for training purposes.

Section 3 — What Is Sent to AI Providers

Sent ✓
✓ Workflow name and description
✓ Individual step titles and descriptions
✓ Systems involved per step
✓ Data fields required per step (names only — no actual values)
✓ Platform selections (e.g. Salesforce, SAP, Epic, Guidewire)
✓ Agent type (e.g. Claims Triage Agent, KYC Onboarding Agent)
✓ Schema metadata you upload:
— Column names
— Data types
— Null counts and row counts
— Table names and relationships
✓ Your company name (report cover page only)
Never Sent ✗
✗ Raw database records or customer data
✗ Personally identifiable information (PII)
✗ Protected health information (PHI)
✗ Financial transaction records
✗ Social Security numbers or national ID numbers
✗ Passwords, API keys, or credentials
✗ Proprietary algorithms or source code
✗ Email content or communications data
✗ Any data with actual field values

Section 4 — Data Lifecycle

  1. You enter workflow description and upload schema metadata
  2. Data transmitted to TekCapitol over HTTPS
  3. Data sent to AI provider over encrypted connection for analysis
  4. AI provider processes and returns findings
  5. AI provider does not retain data after processing your request
  6. TekCapitol generates and delivers your report
  7. Your assessment data retained by TekCapitol for 90 days
  8. Permanently deleted after 90 days

Section 5 — For Regulated Industries

If your workflow involves any of the following, contact us before uploading any data:

· Patient data or clinical workflows — HIPAA
· Financial services or trading — FINRA · SEC
· Debt collection workflows — CFPB · Reg F
· EU customer or employee data — GDPR
· Federal government systems — FedRAMP
· Children's data — COPPA

Email: info@tekcapitol.com
Subject line: "Regulated Industry Assessment"

We will confirm the appropriate data handling approach before you proceed. This does not prevent you from using Kyklos360 — it ensures your specific compliance requirements are met correctly.

Section 6 — Enterprise Data Processing

For enterprise clients requiring a formal Data Processing Agreement (DPA) or vendor security questionnaire response, contact:

info@tekcapitol.com
Subject: "DPA Request" or "Security Questionnaire"

We will respond within 3 business days.

Section 7 — Questions

TekCapitol, Inc.
San Jose, California
info@tekcapitol.com
tekcapitol.com/ai-disclosure