Backend · AI systems
I build production systems end to end: backends, pipelines, automation. The last couple of years mostly around AI: RAG, agentic pipelines, and model evals.
An engineer with seven years of commercial experience. I build production-grade systems end to end: backends, pipelines, automation, from gathering requirements to deploy and support. Background in fintech and high-load.
I've worked both solo and on teams, from startups to large international ones. These days I write less code and focus more on the product side.
For the last couple of years I've worked closely with AI, not just as a tool but from a product angle: building RAG systems, agentic pipelines, and writing evals for models.
Value-wise I'm drawn to education, quality of life, privacy, civil liberties, and AI safety. I contribute a bit to open source I care about, and want to do more.
Author
This very site. After a rework it became a clean open-source template: Astro frontend, FastAPI backend, DB-driven content with an admin panel, two locales, CI/CD with auto-deploy, fully Dockerized. Running in production.
Author
A multi-agent orchestrator on LangGraph: from a task description in Telegram it assembles a working project in production — code, CI/CD, domain, SSL. Agents are graph nodes, work runs in isolated worker containers, generation is spec-first.
Author
A spec-first Copier framework for AI-agent-ready microservices. FastAPI + Postgres with bot/notifications/frontend modules; pull infrastructure updates via copier update without touching your code. Ruff/Mypy/Pytest and CI out of the box.
Author
A text RPG with multi-level LLM world simulation. An LLM dungeon master orchestrates a layered world: geography, weather, settlements, NPC conversations, each layer on top of the last. Python backend, Vite frontend.
Core stack and tooling.
Designing resilient APIs and services ready to scale.
Implement ML/LLM agents, pipelines, and MLOps practices.
Ship CI/CD, monitoring, and infrastructure automation.
“I highly recommend Vladislav for his exceptional work in creating a Telegram bot for our cryptocurrency project. He demonstrated a high level of expertise, professionalism, and attention to detail throughout the project. His deep understanding of Python, the Telegram API, and cryptocurrency exchanges and APIs was critical in creating a reliable bot that provided our users with real-time alerts. His communication and updates were prompt and frequent, and he consistently exceeded our expectations. I highly recommend Vladislav for any project that requires expertise in developing Telegram bots or web parsers.”
“Vlad helped me get to grips with FastAPI, Docker, Redis, and Postgres, and he answered my questions over chat. Trying to figure it all out alone felt overwhelming due to the amount of information; Vlad’s support made a huge difference.”
“A consummate professional who can handle any programming task. He explains things clearly and is a great listener. Willing to take on challenges of any complexity.”
Message me on Telegram @vladislav_meshk or send an email to vladmesh.dev@gmail.com .