Browse categories
Explore
Fiverr Pro
English
$
USD
I will set up claude code ai agents on your vps with docker
Senior DevOps and Network Security Engineer and Linux Expert
Stop worrying about server downtime and security breaches.
Hello, I’m Mete. I am a Senior System & Network Engineer with over 7+ years of experience in designing, securing, and automating high-perfor...
About this Gig
I've been running production Linux fleets for years. Setting up AI agents isn't just "docker run" it's about making them survive reboots, handle crashes, stay secure, and not melt your GPU bill.
What you get:
Dockerized agent Claude Code, OpenHands, AutoGen, or your custom Python/TypeScript agent, multi-stage build, non-root, health checks
Systemd service Starts on boot, restarts on crash, resource limits (CPU/RAM/GPU), log rotation via journald
Nginx reverse proxy TLS (Let's Encrypt auto-renew), WebSocket support for streaming, rate limiting, basic auth option
Redis queue Job queue with priority, retry/backoff, dead-letter handling (Standard/Premium)
Observability Loki + Promtail for logs, Prometheus metrics (queue depth, latency, token usage), Grafana dashboard
Vector DB Qdrant/Weaviate for RAG agents, persistent volume, backup cron (Premium)
Cost guard Daily spend alert (Discord/Slack/Email), auto-shutdown on budget exceed (Premium)
Runbook Architecture, scaling guide, "agent stuck?" troubleshooting, credential rotation procedure
Stack I work daily: Docker, Nginx, systemd, Redis, Loki/Prometheus/Grafana, Qdrant, NVIDIA Container Toolkit, RunPod/Vast.ai/Lambda APIs, Linux
My Portfolio
Other DevOps Engineering Services I Offer
FAQ
Which agents do you support?
Claude Code, OpenHands, AutoGen, CrewAI, LangGraph, custom Python/Node agents. Any containerizable agent works. Message me with your repo — I'll confirm compatibility.
How do you handle API keys / secrets?
Never in image or compose. Options: 1) HashiCorp Vault, 2) 1Password CLI, 3) SOPS/age encrypted files, 4) Docker secrets (swarm), 5) GitHub Environments (CI/CD). We'll pick what fits your workflow.
Can you optimize GPU costs?
Premium only: Daily spend alerts (Slack/Discord/Email), auto-shutdown idle agents, spot instance fallback (RunPod/Vast.ai), token usage tracking per agent. Typical savings: 40-60% vs always-on.
