AI & Digital solutions. Software development. 15+ years of experience.

Hi, I'm Petro Lashyn. I design and build AI and digital solutions for businesses — Graph RAG, Model Context Protocol and agentic architectures — on top of software that lasts: clean architecture, PostgreSQL, DDD and Laravel. CTO at Well Digit.

Get in touch
Petro Lashyn

Expertise

Architecture

Systems should be legible five years from now. I design for maintainability first — clean boundaries, explicit contracts, boring infrastructure. Laravel, DDD. The stack that lets you ship and sleep.

AI Integration

Graph RAG. Model Context Protocol. Agentic orchestration. Building the infrastructure layer that makes LLMs actually useful inside existing systems — not demos, production.

Consulting services

Current Focus

Business software has always had two things: actors and systems. The only bridge was UI — humans adapting to software logic. Now there's a new layer possible. One where users say what they need and the system understands, responds, and acts. Not a chatbot. A genuine intelligence embedded inside your product. I'm building this AI layer using Graph RAG, MCP, and agentic architectures — on top of existing business systems, not replacing them. Currently focused on regulated industries, where AI must be traceable, evidence-backed, and human-supervised. The patterns here aren't established. That's exactly why I'm interested.

Latest writing

All articles
Kalimera: From laravel new to a Production-Ready App in One Run An open-source interactive installer that takes you from laravel new to a fully configured, quality-checked app running in Sail — Pint, PHPStan, Rector, Vet, Horizon, Scout, Boost and an AI-agent guard included — in one guided CLI flow. Laravel 7 min My Uptime Monitoring Setup How I monitor my production systems with Barkme, Laravel Forge, Laravel Nightwatch and a few Telegram alerts. DevOps 4 min Knowledge as Code: Connecting Business Knowledge to AI with Git and MCP Records answer WHAT and WHO. Knowledge answers WHY and HOW. Keep that knowledge in Git as one source of truth, edit it in a simple web editor, and serve it to every AI tool through MCP. AI 6 min Jev as System One Grading code with Jev didn't work. Asking it concrete yes/no questions did. AI 12 min