Architecture Advisory
Make high-impact technology decisions with a clear view of trade-offs and delivery reality.
- Current-state assessment
- Target architecture
- Technology evaluation
- Architecture decision records
- Risk and dependency analysis
Independent consulting for cloud-native platforms, Kubernetes, observability, OpenTelemetry, and applied AI. From strategy and architecture to implementation guidance and technical delivery.
Architecture is most useful when it connects business intent, engineering choices, and the operational system that teams must own.
Architecture assessments, target-state design, decision records, modernization roadmaps, governance models, and stakeholder alignment.
Developer platforms, paved roads, self-service workflows, GitOps delivery models, platform APIs, and operational ownership boundaries.
Kubernetes platform architecture, cloud landing zones, multi-cluster strategy, workload identity, networking, storage, and migration planning.
OpenTelemetry strategy, metrics, logs and traces, collection architecture, scalable backends, alerting design, dashboards, and operational adoption.
AI-assisted software delivery, agent workflows, context engineering, retrieval-augmented systems, guardrails, evaluation, and enterprise adoption patterns.
Delivery architecture, CI/CD, reliability models, operational readiness, release governance, incident learning, and measurable engineering outcomes.
Consulting covers discovery, target architecture, technical design, delivery guardrails, and implementation support—keeping strategic intent connected to what runs in production.

I’m Mihály Petrényi. I work directly with technology leaders and engineering teams, bringing architecture-level perspective close enough to implementation to remain useful.
About my approachFocused engagements shaped around the problem—not a fixed methodology or a stack selected in advance.
Make high-impact technology decisions with a clear view of trade-offs and delivery reality.
Create an operable platform foundation that helps engineering teams move with confidence.
Introduce AI-enabled workflows with useful context, safe tools, and accountable ownership.
Establish context, constraints, risks, and the decisions that matter.
Create a coherent target architecture with explicit trade-offs.
Test assumptions through review, prototypes, and operational scenarios.
Translate decisions into guardrails and support teams through execution.
FleetPilot is an open-source, cloud-native fleet management platform for observability agents, built around the Open Agent Management Protocol.
A work in progress exploring how agent configuration, control, and telemetry can become safer and easier to operate at scale.