Browse categories
Explore
Fiverr Pro
English
$
USD
I will setup fix hermes ai multi agent n8n mcp claude code ollama obsidian memory crm
Roman
About this gig
Imagine an AI that researches competitors, qualifies leads, replies to DMs, and deploys code all before your morning coffee.
I build and configure complete AI agent systems on your VPS Hermes Agent, n8n, Claude Code, Codex, OpenClaw, MCP servers & Ollama self-hosted, secured, and running 24/7
What I set up:
- Build your multi-agent RAG pipeline with LangGraph, CrewAI or Vapi Retell
- Migrate your OpenClaw, Zapier or Make setup to Hermes or n8n
- Wire your Hermes Agent to Telegram, Slack, WhatsApp or Discord
- Connect your n8n to GoHighLevel, HubSpot, Airtable or Apollo Clay
- Configure your Hermes MCP with Composio, GitHub, Notion or Supabase
- Build your Codex or Windsurf multi-agent setup with Claude Code Supabase
- set up your Browserbase or Playwright browser agent with Firecrawl Tavily
- Fix your MCP server failed to start in Claude Desktop, Cursor or Windsurf
- Wire your Odysseus agents to MCP, Playwright, opencode and OpenRouter
- Build your custom MCP server in Python or TypeScript for GitHub, Postgres or HubSpot
- Build your Paperclip agent org with CEO agent, sub-agents, budget caps and audit logs
Ready to automate everything? Send me a message and let's build
Get to know Roman
Roman
AI desktop automation with claude cowork, no coding required
- FromUnited States
- Member sinceJun 2026
Languages
English, French, German
Claude Cowork is one of the most powerful productivity tools available right now and most people barely scratch the surface of what it can do
I set it up for you properly. Your files, tasks, and daily workflows configured to run smoothly in the background no coding, no technical headaches, no half-finished setup collecting dust.
The result? Hours back every week, and a workday that actually feels lighter.
I tailor everything to how you work specifically, and most clients are up and running fast.
If that sounds worth a conversation, let's talk
FAQ
Can you build a custom MCP server and connect it to Claude Desktop or Cursor?
Yes. I build MCP servers in Python and TypeScript and connect them to GitHub, Postgres, Supabase, Notion, Linear, Jira, HubSpot, Firecrawl and Composio. I fix failed to start errors in Claude Desktop, Cursor and Windsurf.
What n8n automations do you build?
I build voice agents with Vapi and Retell, inbound outbound call agents, appointment booking AI, WhatsApp Business API bots, Instagram DM bots, GoHighLevel and HubSpot CRM automation, SEO automation, deep research agents, competitor monitoring, YouTube-to-blog, TikTok and UGC content pipelines.
Can you set up Mem0 so my AI agent remembers between sessions?
Yes. I set up Mem0, OpenMemory MCP, Letta, Zep and Cognee for Hermes, Paperclip, LangGraph and CrewAI. I configure pgvector, Qdrant and Chroma for storage. I verify memory recall before marking the order complete.
Can you set up Vapi or Retell voice AI with GoHighLevel and n8n?
Yes. I set up Vapi and Retell and connect them to GoHighLevel, HubSpot and n8n webhooks. I configure ElevenLabs voice and call transfer logic. I fix Vapi webhook errors and Retell agent not responding issues.
What additional Hermes Agent integrations do you support?
I connect Hermes Agent to Perplexity, You.com, Linkup, SearXNG, Hunter.io, AgentMail, Honcho, Hindsight, Supermemory, Vectorize.io, Milvus 2.6, Higgsfield, Polymarket, Apple Health, Spotify, Tailscale, Zed editor ACP, LI.FI, Pimlico and EIP-7702.
Can you connect my AI agent to Slack, Telegram, WhatsApp and other channels?
Yes. I connect Hermes Agent and Paperclip to Telegram, Slack, Discord, WhatsApp, Signal, Microsoft Teams and Email. I configure Composio, Browserbase, Playwright and Tailscale for secure private access across your full agent stack.
How does Hermes Agent compare to LangChain, LangGraph or AutoGen?
Hermes Agent is lighter, faster to deploy and requires no complex orchestration code. Where LangChain, LangGraph, AutoGen, MetaGPT and CrewAI demand heavy engineering overhead, Hermes gets your agent running on VPS in hours with built-in memory and skills.
Is Hermes Agent better than Manus, Lindy AI or Superagent?
Hermes Agent is fully self-hosted meaning you own your data, Manus, Lindy AI, Superagent and Vellum are managed cloud platforms Hermes gives you the same power at a fraction of the cost.
Can you deploy Odysseus AI on my VPS and expose it securely?
Yes. I deploy Odysseus on Hetzner, DigitalOcean, Hostinger, Contabo and Linode with Docker Compose. I secure it via Tailscale, Cloudflare Tunnel, Cloudflare Access, Caddy, nginx or Traefik with HTTPS and AUTH_ENABLED.
Can you set up Open WebUI, Repello or Ombre?
Yes. I deploy Open WebUI, Repello and Ombre as part of a complete self-hosted AI stack on your VPS alongside Ollama and your agent systems.
