Available for projects
AINFRA
IT Infrastructure & AI Engineering by
ShereOn Group

AINFRA by ShereOn Group

15+ years enterprise IT. Self-hosted AI, LLM pipelines, remote desktop infrastructure, Asterisk telephony, monitoring stacks. No cloud lock-in. Full data sovereignty.

15+
Years experience
54k
Docs indexed (RAG)
3k
VMs managed
7
Service pillars
Explore packages View real cases →
Ollama Dify Weaviate Proxmox Asterisk PBX Prometheus Grafana Zabbix WireGuard OPNsense Vaultwarden n8n Qwen3-32B Mistral 7B NemoClaw OpenClaw Hermes HashiCorp Vault MS Active Directory Docker Ansible GitLab CI/CD Ollama Dify Weaviate Proxmox Asterisk PBX Prometheus Grafana Zabbix WireGuard OPNsense Vaultwarden n8n Qwen3-32B Mistral 7B NemoClaw OpenClaw Hermes HashiCorp Vault MS Active Directory Docker Ansible GitLab CI/CD
What I build

SEVEN
PILLARS

01
AI & LLM Systems

Self-hosted AI assistants, RAG pipelines, vector databases, multi-model setups. Zero cloud dependency.

OllamaDifyWeaviateQwen3-32BRAG
02
Remote Workplaces

Full remote desktop infrastructure from bare metal to end-user login. Proxmox, AD, RDS, VPN.

ProxmoxMS ADRDSOPNsenseWireGuard
03
Monitoring & Alerting

Full observability stack. Metrics, logs, dashboards, alerts to Telegram and Slack.

PrometheusGrafanaZabbixAlertmanager
04
Telephony

Self-hosted PBX with IVR, SIP trunks, call recording. From small offices to enterprise.

AsteriskSIPIVRCDR
05
Self-Hosted Services

Private alternatives to SaaS. Password managers, encrypted sharing, Git, helpdesk — all on your server.

VaultwardenGLPIGiteaPortainer
06
Automation & DevOps

Automate anything. Onboarding/offboarding, Jira workflows, CI/CD pipelines, API integrations.

n8nAnsibleGitLab CIJira
Fixed-scope packages

PICK
YOUR STACK

Every package includes deployment, configuration, documentation and handover. Contact us to get started — we'll respond with a tailored proposal.

Basic
AI Starter
$1,800
one-time setup
  • Ollama + 1 model (Mistral / Qwen3 / Llama)
  • Dify / Flowise / AnythingLLM orchestration
  • Basic RAG — up to 5,000 docs (Chroma)
  • REST API endpoint
  • Docker + systemd deployment
  • Documentation & handover
Get in Touch
Custom
AI Enterprise
From $6,000
scope discussed individually
  • All Pro +
  • Multi-instance / multi-GPU deployment
  • Model selection: DeepSeek-R1, CodeLlama, Mixtral 8x7B, Phi-4, Gemma 2
  • Fine-tuning consultation
  • Multi-tenant architecture
  • CRM / ERP API integration
  • SLA-backed support included
Let's discuss
AI Support
Monthly maintenance — model updates, KB sync, quality monitoring, 8h SLA
$400/mo
Get in Touch
Standard
Remote Office
$1,800
one-time setup
  • Proxmox cluster (up to 3 nodes)
  • MS Active Directory + DNS + DHCP
  • RDS — up to 20 users
  • WireGuard or OpenVPN
  • OPNsense firewall
  • User onboarding/offboarding process
  • GPO basic configuration
  • Documentation & handover
Get in Touch
RDS Support
User management, VPN, updates, performance monitoring — 8h SLA
$600/mo
Get in Touch
Standard
Monitoring Starter
$800
one-time setup
  • 1 system: Prometheus+Grafana OR Zabbix
  • Up to 20 services / hosts
  • 1 team
  • 1 messenger: Telegram OR Slack
  • Basic dashboards (CPU, RAM, Disk, Network)
  • 15+ alert rules
  • Documentation
Get in Touch
Monitoring Support
Alert triage, new services, dashboard updates — 4h SLA for critical
$300/mo
Book a call
Standard
PBX Starter
$1,100
one-time setup
  • Asterisk PBX self-hosted
  • SIP trunk (up to 2 providers)
  • Inbound / outbound routing
  • Basic IVR (up to 3 levels)
  • Up to 20 extensions
  • IP phones setup (Grandstream / Cisco / Yealink)
  • Documentation
Get in Touch
Telephony Support
Extensions management, SIP troubleshooting, IVR changes — 8h SLA
$350/mo
Book a call
Service What it is Setup Support/mo
VaultwardenSelf-hosted Bitwarden — team password manager$400$100
PrivateBinE2E encrypted paste — zero-knowledge sharing$250$60
GiteaSelf-hosted Git — your own GitHub$350$80
PortainerDocker management UI — no CLI needed$300$70
GLPIIT asset management + helpdesk ticketing$600$150
PassboltOpen-source team password manager$400$100
NextcloudSelf-hosted file storage + collaboration$500$120
Uptime KumaSelf-hosted status page + uptime monitoring$200$50
Stack Deal — the more you deploy, the less you pay for support
2 services → 2nd support at -50% · 3 services → 3rd support FREE · 4+ services → pay for 2, get all
Small
Task Automation
From $300
single task, one tool
  • Slack / Telegram reminders & alerts
  • Data sync between 2 systems
  • Daily / weekly automated reports
  • Auto Jira task creation from triggers
Get in Touch
Large
Full Pipeline
From $1,500
full workflow, multiple teams
  • Full onboarding pipeline: PeopleForce + Jira + AD + Slack + Confluence
  • GitLab CI/CD pipeline setup
  • Jira Advanced + JQL + reporting
  • Cross-team coordination automation
  • All tools discussed individually
Let's discuss

Tools: n8n (self-hosted) · Make · Zapier · Activepieces · Ansible · Bash/Python · GitLab CI/CD · Node-RED

Basic
Infrastructure Support
$500/mo
up to 10 servers
  • Availability monitoring
  • Monthly updates & patches
  • Backup verification
  • 8h SLA (business hours)
  • Monthly report
Get in Touch
Enterprise
Enterprise Support
From $2,000/mo
scope discussed
  • All Pro +
  • Dedicated support
  • 1h SLA for critical incidents
  • 24/7 proactive monitoring
  • Weekly reports
  • Custom SLA terms
Let's discuss
Real work

PRODUCTION
DEPLOYMENTS

✓ Production · AI

Nemo AI — Travel Platform Support

B2C/B2B travel booking platform (confidential)

Full self-hosted RAG system on bare-metal GPU server. Polls 5 HelpScout mailboxes every 2 minutes, filters automated noise (~60% of tickets), extracts customer identity across 6 languages, runs RAG analysis against 54k-doc knowledge base, generates formal reply drafts, delivers to Slack — all within 2 minutes. Zero cloud APIs.

OllamaDify 1.13Weaviate PostgreSQLRedisPrometheus GrafanaPythonHelpScout APISlack Block Kit
54k
Docs indexed
2min
Response time
60%
Noise filtered
6
Languages
✓ Production · Infrastructure

Enterprise Remote Desktop — SoftServe + DTEK

SoftServe Poland + DTEK — 3,000+ VMs managed

Full lifecycle management of VMware/Hyper-V/Azure clusters. MS AD, RDS farms, SCCM/SCVMM/SCOM. Proxmox virtualization, Zabbix + CheckMK monitoring, Avamar backup, Fortinet perimeter, HashiCorp Vault secrets management.

ProxmoxVMwareHyper-V MS ADRDSZabbixSCCM
3k
VMs managed
99.9%
Uptime target
✓ Production · Telephony

Asterisk PBX + Full Self-Hosted Stack

Bookimed / AllBiz — telephony & internal services

Asterisk PBX with IVR, SIP trunks, call recording, CDR reporting. Zimbra mail server, OpenFire XMPP chat, OTRS helpdesk, Zabbix + Grafana monitoring. Ansible automated deployment on-premises Linux stack.

AsteriskZimbraOpenFire OTRSZabbixGrafanaAnsible
8+
Services deployed
100%
On-premises
How I work

SIMPLE.
FIXED SCOPE.

01
Get in Touch

Contact us via the form or AI phone line. We'll gather your requirements, understand your setup, and prepare a tailored proposal — no commitment required.

02
Fixed-Scope Quote

Clear proposal with exact scope, timeline, and price. No hourly billing surprises. 50% upfront, 50% on delivery.

03
Deploy & Handover

I deploy, configure, test, and document everything. You get full access, complete documentation, and a working system.

FAQ
STILL
HAVE
QUESTIONS?

Can\'t find the answer here? Contact us via the form or AI phone line and we'll get back to you promptly.

Get in Touch
Proxmox, VMware, Hyper-V, MS AD, RDS, OPNsense, WireGuard VPN, Asterisk PBX, self-hosted AI stacks (Ollama, Dify, Weaviate), monitoring (Prometheus, Grafana, Zabbix), and self-hosted services (Vaultwarden, GLPI, Gitea). Full list in the packages section.
Yes, 100% remote. I'm based in Split, Croatia and work with clients worldwide. All deployments are done via SSH, VPN, or remote management tools.
Starter packages — 1-2 weeks. Pro packages — 2-4 weeks. Full Stack + AI — 4-8 weeks. Timeline is agreed and fixed before we start. No surprises.
That's exactly the point. Everything runs on your server. No third-party APIs, no cloud lock-in, no data leaving your infrastructure. You own the data, the models, and the system.
Yes — monthly retainer available for every service. Includes updates, monitoring, alert triage, and SLA-backed response. Details are in each package section above.
Minimum for Mistral 7B: 16GB RAM, 4-core CPU, 50GB SSD. Recommended for Qwen3-32B: 64GB RAM, NVIDIA GPU with 8GB+ VRAM. I can advise on hardware selection during the initial consultation.
Let's talk

READY TO
BUILD YOUR
STACK?

Reach out via the form or call our AI agent directly — we'll gather your details, answer questions about our services, and register your inquiry. Our team will get back to you with a tailored proposal as soon as possible.

Send a message