I am looking for an experienced AI/Linux engineer to create a self-hosted AI infrastructure on my server.

Budget: 27000 UAH Deadline: 14 days

Based on experience - we have created self-hosted AI infrastructure, agent workflows, corporate automation, and systems with integrations. For the stack, I would choose Ubuntu Server or Debian, Docker Compose, Ollama, Open WebUI, Telegram Bot API, Postgres, Qdrant or Chroma, Nginx, Tailscale or WireGuard, NVIDIA Container Toolkit, Prometheus, and Grafana.

Regarding cost - 45,000 UAH for the initial phase lasting 10-14 days. This includes basic architecture, installation, GPU acceleration, access, monitoring, memory, Telegram and Web UI, a foundation for agents, and documentation for further development. A budget of 1,000 UAH here will likely not even cover proper server diagnostics =/

Note, there is a nuance - such a system is better built not as a set of scripts, but as a modular platform. First, we establish a stable core, then add new workflows, agents, and integrations without redesigning the entire architecture.

I would like to clarify 2 points:
> What graphics card, how much VRAM, RAM, and disk space on the server?
> Are external accesses needed for users or only closed access for you?
Similar cases:
> https://business.ingello.com/vorfahr - AI and automation of business processes
> https://business.ingello.com/fractal - agent approach and modular logic of automation
> https://business.ingello.com/forma-bpm - workflows, roles, and managed automation

Main landing page for Ingello for FLH - https://systems-fl.ingello.com

!!Important!! I would not start with a chaotic installation of everything at once. Measure seven times - cut once. First, a brief design phase, then installation and security checks; otherwise, the system will quickly become an expensive toy rather than a working tool. Expand

Budget: 1000 UAH Deadline: 30 days

I will deploy a stable and fault-tolerant self-hosted AI ecosystem based on Ubuntu Server with GPU acceleration, isolated Docker containers, and secure remote access via Telegram.

Which orchestration platform do you plan to use as the core for managing multi-agent scenarios — the visual n8n, CrewAI, or custom code in Python?

I am ready to discuss the budget, exact timelines, and stack for long-term collaboration in detail with you in private correspondence.

Budget: 1000 UAH Deadline: 7 days

Hello, I have worked on AI infrastructure for automating business processes for a company with 50+ employees - I set up Ollama with local LLMs, created a multi-agent system with Telegram integration, achieving a 40% reduction in manual work.

I’m curious if you plan to integrate the system with existing corporate applications, or will it be a standalone solution for internal processes?

I suggest we get in touch; I will provide you with free technical consultation and we can outline a development plan + I will tell you about my team! ✨

Budget: 27000 UAH Deadline: 14 days

We already have a nearly ready database for such AI infrastructure, which can be quickly adapted to your server and launch the first working version. ))

In terms of cost, I would estimate starting from 45,000 UAH for the first stage over 10-14 days.

This includes a Linux server, Docker ecosystem, Ollama, Open WebUI, Telegram integration, basic memory, remote access, monitoring, launching local models, and a foundation for agents.

1,000 UAH seems suitable only for a short server diagnosis or consultation, not for building a stable system.

For the stack, I would recommend Ubuntu Server, Docker Compose, Ollama, Open WebUI, PostgreSQL or SQLite for initial memory, n8n or a custom workflow layer, Telegram bot API, Prometheus or Netdata for monitoring, NVIDIA drivers and CUDA if a suitable graphics card is available.
We have done similar solutions for AI agents and automation.

https://business.ingello.com/fractal - closely related in logic to multi-agent processes and architectural development.

https://business.ingello.com/vorfahr - an example of automation with AI and a systematic approach to the product.

https://business.ingello.com/tts - an example of AI integrations and working with application scenarios.

The main page for our system projects for Freelancehunt - https://systems-fl.ingello.com

For access, we need the server parameters, the presence or model of the NVIDIA GPU, the desired method of remote access, and a list of the first 3-5 tasks that the assistant should automate.

I would like to clarify two points.

The server has already been purchased and there is physical access or IPMI.

The assistant should only help through Telegram/Web UI or also run commands, create files, and change workflows on the server.

Overall, it’s fine to go step by step - first a stable core, then agents, memory, and new working scenarios.

We won’t build a tower to the heavens, but it’s better to lay the foundation correctly =/-- Expand

Budget: 16000 UAH Deadline: 10 days

Good day! I have extensive experience in deploying such infrastructure in Kubernetes. Architecturally, it might be more appropriate to use vLLM. I can start working soon with a preliminary meeting to agree on the details. I look forward to your message.

Budget: 27000 UAH Deadline: 10 days

Good day!

A week ago, I took 3rd place solo at the AI Agent Olympics Hackathon (Milan AI Week 2026, the largest AI event in Europe) out of 731 teams - built a multi-agent system in 7 days. Over a year as a full-time AI automation engineer, prior to that 4 years as a PM in the US/EU/Asia, MSc in Strategic PM, PRINCE2, Python/Docker daily.

Stack for May 2026: Ubuntu 26.04 LTS, Docker, Ollama v0.6.2 (Llama 4 Scout / DeepSeek V4 / Qwen 3.6 for your hardware), Open WebUI v0.8.0 (with patch CVE-2026-0765), n8n self-hosted as orchestrator, Supabase pgvector for memory, Telegram bot with human-in-the-loop, CUDA 13.2.1, Tailscale + Prometheus/Grafana.

MVP (installation + 2-3 agents + Telegram + monitoring): 35,000 UAH, 10-14 days. Full modular ecosystem with auto-generating agents: 70,000 UAH, 3-4 weeks. Let me know what hardware we are deploying on - I will return a precise estimate and roadmap within a day.

Cases in the profile.

Budget: 1000 UAH Deadline: 1 day

Good day!

I have experience in building self-hosted AI infrastructure on Linux with Docker, GPU, Telegram integration, and automation systems.

For your project, I would recommend using vLLM instead of Ollama as a more efficient and scalable solution for local LLMs, as well as Open WebUI, PostgreSQL/Redis, and an agent system on LangGraph or CrewAI.

The preliminary timeline for implementing the basic platform is 2 weeks, and the cost depends on the number of agents, automation scenarios, and autonomy requirements. I am ready to discuss the details and propose an optimal architecture for further scaling.

Budget: 27000 UAH Deadline: 30 days

Hello. The task is clear: you need not just an "installed Ollama," but a full-fledged Self-hosted Autonomous AI Ecosystem with agent logic, memory, and scalability. I have 8+ years of experience in developing complex systems and deep expertise in AI infrastructure (local LLMs, RAG, Multi-agent systems).

My recommended stack:

Infrastructure: Ubuntu Server + Docker Compose + Portainer (management) + NVIDIA Container Toolkit.
LLM Core: Ollama (quick start) or vLLM (high performance) + Open WebUI with Multi-user/RAG support.
Workflow & Agents: n8n (self-hosted) — the ideal "glue" for automations + LangGraph or CrewAI for complex multi-agent logic.
Memory/RAG: PostgreSQL with pgvector extension or Qdrant for long-term agent memory.
Access & Sec: Tailscale/Netbird (secure remote access without opening ports) + Uptime Kuma (monitoring).
Implementation stages:

Base Setup (3–5 days): OS, GPU drivers, Docker, Ollama, Open WebUI. Setting up secure access.
Agent Logic (5–10 days): Deploying n8n, creating a Telegram bot, setting up the first autonomous workflows (for example: data collection -> AI analysis -> report in TG).
Advanced Architecture (7–10 days): Integrating Multi-agent framework (CrewAI/LangGraph) to perform complex cross-functional tasks and building a RAG system (memory).

Cost: Ecosystem deployment — $1,000. Further creation of complex agents and workflows — by agreement (hourly or per project). Timeline: 2–4 weeks until the full launch of the autonomous system.

Clarifications:

What specific GPUs are installed on the server (VRAM is critical for model selection)?
What are the top 2–3 tasks the assistant should automate immediately after launch? Expand

Budget: 15000 UAH Deadline: 10 days

Добрий день!

Маю досвід із Ubuntu Server + Docker Compose, Ollama + Open WebUI, n8n workflows, PostgreSQL pgvector, Telegram bot (aiogram), Tailscale. Налаштовував NVIDIA Container Toolkit для GPU-прискорення.

Уточніть: яка GPU і скільки VRAM? n8n чи LangGraph для workflows? Це вплине на терміни.

Орієнтовна вартість повного setup: 15 000–20 000 UAH, 10–14 днів.

Budget: 8000 UAH Deadline: 1 day

Good day. I am ready to complete this project and have extensive experience in developing various applications.

Budget: 10000 UAH Deadline: 6 days

Hello, Vebster, it’s quite strange to implement such a system from scratch now, considering the huge number of ready-made open-source solutions. I suggest going through the following:
1. The classic that started it all: OpenClaw
2. A smarter and more adaptive solution: Hermes Agent
3. The latest, freshest, and most reliable solution: OpenHuman

It has practically everything you need and much more.
If you are against such autonomous agents, we can build from scratch:
1. Better to use llama cpp instead of ollama - faster and more efficient (in benchmarks, up to 30% faster token generation)
2. PydanticAI or LlamaIndex are good for agents
3. There are many approaches for memory, currently, RAG + obsidian graph combinations show good results I have already created similar systems multiple times. I have built multi-agents for programming and project evaluation. Also, my last project relates to the use of agents. So, I have experience.
If we use a ready-made agent, the price and deadlines will be lower; I will indicate the rate for the second option.
I would be happy to work with you! Expand

Budget: 1000 UAH Deadline: 3 days

Hello! Your project looks very promising. I am ready to start working on it and complete it at the highest level.

Budget: 27000 UAH Deadline: 14 days

Hello. I have experience with Python, Docker, FastAPI, Telegram bots, self-hosted AI infrastructure, RAG/memory, local LLMs via Ollama, and building automation workflows.

For the stack, I see it like this: Ubuntu Server, Docker Compose, Ollama + Open WebUI, Qdrant/PostgreSQL for memory, Telegram bot as the interface, n8n or a custom workflow layer, monitoring, backup, and secure remote access.

I would approach it step by step: first a stable core on the server, then Telegram/Web UI, memory, agents, workflow automation, and GPU acceleration. This way, the system won't turn into an "architectural mess" and can be scaled.

I am ready to discuss the tasks and propose a realistic architecture for your server.

Budget: 1000 UAH Deadline: 7 days

Привіт!

Профільний стек — крутю Ollama + Open WebUI на своєму воркстейшені (GTX 1660 SUPER, CUDA 13.1), будувала AI Twin agent з Telegram bridge і persistent memory (Postgres + pgvector). Multi-agent через LangGraph — є робочий PoC.

Рекомендую: Ubuntu 24.04 LTS + Docker Compose, Ollama + Open WebUI, n8n або LangGraph для workflows, Postgres + pgvector для memory, Tailscale для remote, Uptime Kuma + Loki/Grafana для monitoring, Telegram bot на aiogram.

Питання до старту:
— Сервер з GPU вже стоїть, чи треба підбирати залізо? Яка модель GPU і скільки VRAM?
— Які моделі планується крутити (Llama 3.1 8B/70B, Qwen, Mistral 7B)?
— Workflows на no-code (n8n/Flowise), чи кастомний Python orchestrator (LangGraph/CrewAI)?
Терміни: базовий setup (Docker + Ollama + WebUI + Telegram + 1 агент + базовий monitoring) — 7-10 днів part-time. Multi-agent з handoff, tool calling і повний monitoring — ще 1.5-2 тижні.

Бюджет 1000 UAH — мабуть placeholder, реалістична fixed-ціна за повний setup з документацією і handover від 15 000 UAH, або погодинно 600 грн/год.

Long-term ок, я працюю part-time стабільно. Готова розписати детальний кошторис після відповідей на питання вище. Expand