Browse categories
Explore
Fiverr Pro
English
$
USD
I will setup and secure your openclaw or hermes ai agent on vps or local
Arnab Modak
About this gig
Most AI agent installs break within a week. Wrong permissions, exposed
ports, no process supervision, secrets sitting in plaintext, no logs. It
works on the demo call and dies on day three.
I set up OpenClaw and Hermes agents that survive production.
What you get:
- Clean install on your VPS, cloud server, or local machine
- Security hardening: user isolation, firewall rules, SSH hardening, and
secrets in an encrypted store
- Process supervision, so the agent restarts itself if it crashes
- Integration with your tools - Gmail, Slack, Telegram, Google Workspace,
CRMs and any REST API
- Model provider setup, including local models if you need privacy
- A written handover document: what was installed, where, and how to
operate and update it
Why me:
- 8+ years engineering, 4 of them as CTO leading a development team
- Lead at an AI and cloud engineering studio
- I run these agents daily on my own infrastructure, not from a tutorial
- WordCamp Speaker
Not sure what you need? Message me before ordering and I will tell you
honestly whether it is possible, what it costs, and what I would do.
Get to know Arnab Modak
Arnab Modak
AI Developer and Full Stack Engineer
- FromIndia
- Member sinceApr 2026
- Avg. response time1 hour
Languages
English, Hindi
AI Developer and Full Stack Engineer. 8 years building production software.
I build AI agents and the full stack around them - OpenClaw and Hermes
deployments, automation pipelines, web apps, APIs, Web automations and cloud infrastructure.
Most AI agent installs break within a week: wrong permissions, exposed
ports, no monitoring. I build them to survive production, not just a demo.
WordCamp Speaker. Clean code you own completely.
Message me with what you need - I will tell you honestly if I can
build it.
My Portfolio
FAQ
Which is better — OpenClaw or Hermes?
It depends on your use case, team size, and where you want the agent to run. Tell me what you need it to do and I will recommend one honestly — including if you do not need a custom setup at all.
Do I need my own server?
For anything that needs to stay online, yes — a small VPS is sufficient, and I will recommend the right specifications for your workload. For personal or occasional use, a local machine works perfectly well.
Will this work on Windows, Mac, or Linux?
Yes, all three. Linux is the best choice for always-on deployments. Windows and macOS work well for local use. I configure the environment appropriately for your platform rather than forcing one setup on everyone.
Do you get access to my data?
I work on your machine, using a user account you create. I do not copy any data off it, secrets are stored in an encrypted store rather than plaintext, and you can revoke my access the moment the handover is complete.
Which model providers do you support?
OpenAI, Anthropic, OpenRouter, and Google, plus fully local models via Ollama or llama.cpp. Bring your own API key, or I can set you up with a local model if privacy is the priority.
Can you connect it to my own internal API?
Yes, provided it exposes an API. If it does not, tell me what system it is and I will tell you exactly what is and is not possible before you order — rather than discovering it halfway through.
