Browse categories
English
$
USD
Connect to your AI
I will deploy a private self hosted ai chatbot on your own server
Jesus Baena
About this gig
Your data never leaves your infrastructure. I deploy open-source LLMs on your own server - no OpenAI, no third-party API, no data leaving your building.
If you handle client files, medical records, legal documents, financial data or anything under GDPR, sending it to a US API is a problem your compliance team already knows about. Self-hosting solves it properly.
WHAT I DO
- Install and configure Ollama or vLLM on your server, VPS or GPU box
- - Pick the right open-weight model for your hardware
- - Optional: a private RAG system so the model can answer questions about your own documents
- - Optional: fine-tune a model on your data, on your machine
- - Web chat interface, API endpoint, or both
- - Full documentation so your team can run it without me
WHY ME
15+ years in operations, including conflict zones where data protection is not a checkbox. I run a five-node production cluster with local inference. I have deployed exactly this for organisations that legally cannot use cloud AI.
WHAT I NEED FROM YOU
Server access (or I will spec what to buy) and a rough idea of what you want the AI to do.
Message me before ordering and I will tell you honestly if your hardware is enough.
Get to know Jesus Baena
Jesus Baena
AI, Automation and Data Systems Consultant
- FromSpain
- Member sinceJul 2026
Languages
Spanish, English, French
I build practical AI and automation systems for data-driven organizations. With 15+ years of humanitarian field experience, including Ukraine, I turn messy operations into reliable tools.
I design and ship RAG chatbots that answer from your own documents, n8n workflow automations that cut repetitive manual work, interactive dashboards and BI reporting, and geospatial mapping and data pipelines (PostGIS, GIS).
I favor open-source, self-hostable solutions you fully own, with clear docs so your team can run them. Have a data or process problem? Let's scope it and build it.
My Portfolio
FAQ
Will it make things up?
It answers from your documents and cites sources; out-of-scope questions get a clear 'not in the provided material' instead of a guess.
Do I have to use a cloud AI provider?
No. The Premium package can run fully self-hosted on your own server, so nothing leaves your infrastructure.
What do you need from me?
Your documents (or access to them) and one example of a good answer, so I can match the tone and level of detail you expect.
Will my data really never leave my server?
Correct. The model weights are downloaded once, then everything runs locally. No API calls out, no telemetry. I can show you the network configuration so your IT team can verify it themselves. This is the entire point of the service.
What hardware do I need?
It depends on model size. A 7-8B parameter model runs usably on a modern CPU with 32GB RAM, better on any GPU with 12GB+ VRAM. Larger models need 24GB+. Message me with your specs before ordering and I will tell you what is realistic, free.
Is a local model as good as ChatGPT?
For general open-ended chat, no - not yet. For answering questions about your own documents, summarising, extraction and classification, current open models are genuinely competitive. If your use case needs frontier-model reasoning I will tell you before you order.
Can you work with my existing server?
Usually yes - Linux with Docker is ideal. I also work with Proxmox, bare metal and most VPS providers. If your setup will not support it, I will say so before you order rather than after.
Who owns the result?
You do, completely. Open-weight models, open-source tooling, no licence tied to me. The Comprehensive package includes documentation so your team can maintain it without me.
