I'm Vedran Krizic. I bring AI coding agents, clear guardrails and solid architecture into real engineering teams. In my spare time I build open source tools for exactly that.
As a Senior Software Consultant I work at the intersection of software architecture and AI engineering, with enterprise teams across travel, finance, energy and healthcare.
My focus: helping engineering teams adopt AI-assisted software development safely and productively, and designing modern, high-performance software platforms.
Introducing AI coding assistants and agentic workflows into real delivery teams.
Defining instructions, policies and review gates so AI-generated code stays secure and maintainable.
Hands-on AI trainings and coaching for developers, tech leads and architects.
Designing scalable web platforms, from frontend architecture to cloud deployment.
Where AI meets solid engineering: the capabilities I bring to client projects.
The principles I bring to every team, codebase and conversation.
Technology translated into clear business language. If I cannot explain it simply, I am not done thinking.
GDPR, accessibility and legal requirements are built in from day one, not bolted on later.
I would rather make one system clearly better than ship ten forgettable features.
Early architecture decisions make or break a product. I design for growth without rebuilding everything later.
From the server that runs a service to the model that makes it smart: no black boxes, no magic tricks.
Documented, understandable solutions and the knowledge to run them. No lock-in, no secret sauce.
Side projects where I explore agentic development, AI guardrails and modern web platforms.
A bilingual, headless portfolio platform built on Next.js and Strapi 5.
The open source monorepo that powers this very website: a Next.js frontend rendering CMS-driven page sections, a Strapi 5 backend with EN/DE localisation, and a zero-dependency media pipeline that stores uploads directly in SQLite.
A coordinated team of specialised AI coding agents for real-world repositories.
A curated set of agent definitions, skills and coordination protocols for opencode: an orchestrator that plans and delegates, specialists for backend, frontend, performance and documentation, and reviewers that approve or block with evidence.
Policy and instruction templates that keep AI coding assistants on track.
A starter kit for teams adopting AI coding assistants: versioned instruction files, policy templates, a linter that checks them for contradictions and a CI step that flags risky AI-generated changes for human review.
Notes on AI engineering, architecture and the projects above.
Why GDPR, accessibility and a correct imprint belong in the architecture from day one, and how this site handles them.
Roles, contracts and review verdicts: what changed when I stopped prompting one agent and started running a team.
How typed page sections, locale-aware slugs and a strict content contract keep a two-language site maintainable.
Questions about AI-assisted engineering, an invitation to speak, or simply professional exchange: write to me directly.