Insights

Practical views on architecture, production and security.

Krisova publishes concise notes from active architecture, security and production-readiness work. The library will grow progressively.

Krisova method

The Krisova 90-Day Production Path

An architecture-led operating horizon that helps suitable products, decisions, teams, ownership, security, and production obligations converge within a focused 90-day period.

Read method

AI security method

AEPVL: AI Security Production Readiness

A practical architecture method for examining authorization, runtime enforcement, evidence preservation, independent validation, and decision traceability across AI and agentic AI systems.

Read method

Architecture decisions

The Decisions That Become Expensive to Reverse

Platform, data, integration, security, and ownership choices that should be challenged before scale makes them difficult to change.

Coming next

Future security

Preparing Architecture for Post-Quantum Change

A practical view of privacy-preserving, distributed, and quantum-safe readiness without premature technology replacement.

Future brief

AEPVL method

AI security production readiness

AEPVL identifies and validates threat vectors across authorization, enforcement, preservation, validation and logging. It is currently used in architecture design and implementation planning.

AuthorizeWho and what may initiate, access data, use tools or request actions.
EnforceWhere policy becomes runtime behaviour across data, tools and actions.
PreserveWhich context, approvals, versions and evidence must be retained.
ValidateHow outputs, actions and approvals are checked before use.
LogWhat decision trail is required for reporting, audit and review.

Read the AEPVL method