Browse categories
Explore
Fiverr Pro
English
$
USD
I will deploy private self hosted ai on your server with ollama and open webui
Michael B
About this gig
Your team wants ChatGPT-style AI. Your compliance policy says client data cannot leave your control. This gig solves both.
I deploy a private AI stack - Ollama + Open WebUI - on your server or VPS. Prompts, chat history, and documents never touch a third-party cloud. Built for legal, medical, and any business with a no-cloud-AI policy.
Every tier includes:
- Production Docker deployment with HTTPS reverse proxy
- 2-3 open models selected and tuned to your hardware
- Written admin guide: updates, backups, user management
- You own everything: configs, code, admin credentials
- Direct US-based communication
Standard adds RAG - your AI answers from your own documents - plus Tailscale secure remote access for your team. Premium adds an OpenAI-compatible API endpoint for your other software, user accounts, a handover call, and 14 days of support.
About me: B.Sc. Electrical Engineering, ten years in the semiconductor industry. I run this stack in production 24/7 on infrastructure I built and maintain - self-hosted LLMs, Docker services, VPN mesh networking, SIEM logging.
I'm not the cheapest listing here. I'm the one who documents the work and hands you the keys.
Get to know Michael B
Michael B
AI Automation Engineer, Private LLMs, Alerting, Network Security
- FromUnited States
- Member sinceJan 2026
- Avg. response time1 hour
Languages
English
Electrical engineer in Portland, OR (B.Sc., 10 years in the semiconductor industry) who builds AI automation that runs unattended: 40+ scheduled Claude-powered report and monitoring pipelines in production, an LLM that triages my SIEM alerts and pages me only when it matters, and a self-hosted LLM stack (Open WebUI, vLLM, RAG) in daily use. All of it runs on a network I engineered myself - pfSense multi-WAN, VLAN segmentation, WireGuard/Tailscale, Graylog. New to freelancing, not to the work: senior-level engineering at starter rates, documented so you can run it without me.
Other AI Development Services I Offer
FAQ
Why should we trust you with admin access during setup?
I can work over screen share so you watch every step. You issue temporary credentials and rotate everything at handover; the docs show exactly what runs where. All software is open source and nothing phones home to me. NDA on request - and I can demo my own live instance before you order.
Are there ongoing API or licensing costs after delivery?
No. Ollama, Open WebUI, and LiteLLM are open source, and the models run on your own hardware. There are no per-token fees and no subscriptions - your only recurring cost is the server itself (power or VPS rental).
What do I need to provide?
A Linux server or VPS (I can recommend one sized to your budget), a way to grant me access (SSH key or temporary credentials), and a domain name if you want HTTPS on a public address. For Standard/Premium RAG, the documents you want the AI to know.
What hardware does this need?
Open WebUI runs almost anywhere; model speed depends on RAM and GPU. A modern CPU with 16 GB RAM handles small models well; a GPU makes larger models fast. Send me your specs (or budget) before ordering and I will tell you honestly what performance to expect.
What happens after delivery?
Every tier ships with a written admin guide covering updates, backups, and user management. Premium includes a live handover call and 14 days of support. Ongoing maintenance or later upgrades - new models, more users, API integrations - are available as custom offers.
Will it be as good as ChatGPT?
For drafting, summarizing, and Q&A over your own documents, current open models (Llama, Qwen, Mistral) do excellent work. Frontier cloud models are still stronger at hard reasoning - the tradeoff you are buying is total data privacy. I match models to your use case honestly.
