Julien Ziolkowski

AI Automation Engineer · Forward-Deployed Engineer

I turn slow, manual workflows into AI automations that run on your own hardware or via 3rd party integrations (Anthropic etc.) — agents, integrations, and internal tools.

Deutsch

01Profile

Forward-Deployed Engineer, comfortable from the backend to the frontend. I've spent 3+ years across the whole AI stack — down at the frameworks (PyTorch, tinygrad), up through agent harnesses, deployment, and the business side of scoping what's worth building. The work has shipped to real clients, not just demos, and most of it runs self-hosted on hardened VPS boxes or Vercel.

02Experience

Communication Associate
beyondesign Web Design Agency
May 2022 — July 2022
  • Onboarding phase, client-side comms, copywriting, and discovery calls.
AI & Automation Engineer
beyondesign Web Design Agency
July 2022 — until Sep 2026

Prev. Communications Engineer

  • Built and deployed the team's in-house fullstack analysis platform in Go (~15k LOC, single binary, DuckDB, LLM orchestration) that automated per-client review & shop analysis with a custom pipeline. Took a 30-hour manual cycle down to about 10 hours end-to-end.
  • Developed an internal tooling stack & a design system specifically to be AI-Agent friendly.
  • Ran technical discovery and onboarding across 30+ review-analysis projects and 10 data-analysis engagements.
  • Defined measurable e-commerce KPIs and built the systems to nudge them.
  • Wrote custom MCP servers in Go to integrate frontier LLMs with our proprietary data.
  • Set up agentic workflows on VPS behind WireGuard for secure remote team work.
  • Ran hybrid model orchestration - local llama.cpp plus Claude - routed by cost and data sensitivity.
  • Co-built production systems with AI agents against specs I wrote such that I actually define the architecture without creating Spaghetti-code all over the place.
  • Led client calls and walked them through the data.

03Skills

AI & Agent Orchestration
  • Agentic workflows on hardened VPS
  • Routing local (llama.cpp) and frontier models by cost and data sensitivity
  • MCP servers in Go
  • Working ML fundamentals (PyTorch, tinygrad)
  • Building alongside coding agents against my own specs
Sovereign Infrastructure
  • WireGuard-secured private environments
  • Self-hosted, privacy-first deployments
  • Data sovereignty when PII or GDPR is in play
  • Single-binary Go deploys
Full-Stack Engineering
  • Go backends and frontends
  • Reusable, design-system-driven UI components
  • Internal tools and dashboards
  • Deterministic data pipelines
Business Strategy
  • Technical discovery and client onboarding
  • Turning vague goals into measurable KPIs
  • Explaining the technical side to non-techies