Data security
We design so sensitive data never leaves your trust zone. On-premise, private cloud or hybrid with controlled data flows.

ONPREMA.ai guides companies through the full AI implementation lifecycle: from use-case discovery and business case, through building AI agents, RAG knowledge assistants and integrations, to on-premise infrastructure, data security, EU AI Act compliance and support. Private cloud or hybrid — you decide.
Most AI projects stall at PoC. We deliver AI implementation services to production — integrated with your systems, secure and measurable.
We design so sensitive data never leaves your trust zone. On-premise, private cloud or hybrid with controlled data flows.
Every project starts with KPIs and a business case. You know what you save and when it pays back.
GDPR, EU AI Act, NIS2, DORA, sector regulators. We help classify AI systems and prepare documentation and AI governance policies.
We work with open (Bielik, Llama, Qwen, Mistral) and commercial models. Swap models without rewriting the app.
Ready connectors for Mint Service Desk and Apache SOLR, plus experience integrating ERP, CRM, DMS and domain systems.
Consulting, development, infrastructure, training and SLA support — no finger-pointing between vendors.
From AI strategy and readiness audit, through AI agents and RAG deployment, to on-premise infrastructure and training. Pick a single service or a complete AI transformation program.
We map processes, data and systems, identify the highest-ROI use cases and build an AI roadmap tailored to your organization.
Private chat grounded in company documents and semantic search — including on Apache SOLR (hybrid BM25 + vectors) — with source citations.
Agents that run multi-step tasks: ticket handling, document processing, data extraction from invoices and contracts, reporting, sales support.
Native Mint SD plugin (mintsd.com): ticket classification and routing, suggested replies, thread summaries, KB article generation.
GPU sizing and setup, model serving (vLLM, TGI), Kubernetes, monitoring and security. Hybrid architecture with data control where it fits.
We tune models to your domain, jargon and processes and set up evaluation and model lifecycle. The result is your IP.
AI Act system classification, risk assessment, guardrails, protection against prompt injection and data leakage, company AI usage policies.
We train business users and IT teams, then run the solution under an SLA or Managed Service model.
A proven stack we adapt to your environment — bare metal, VMware, OpenShift, Proxmox, private cloud or hybrid.
Workshops, process and data analysis, use-case selection, business case and compliance requirements.
A working solution on your data and infrastructure. KPIs agreed upfront.
Integrations, SSO, hardening, monitoring, AI Act documentation and user training.
More use cases, cost optimization, model updates, SLA or Managed Service.
Three paths — from a quick assessment to a full AI program across your organization.
1-2 weeks
A fast assessment of AI potential in your company.
4-8 weeks
A working solution for one use case on your data.
3+ months
End-to-end AI transformation with support and evolution.
Pricing is always tailored — it depends on scope, number of use cases and infrastructure.
ONPREMA.ai is an initiative by ITSM Software S.A. — a Polish software vendor with a track record of building enterprise-grade systems (including Mint Service Desk, used across Poland and abroad). We combine years of delivery, integration and production support with strong generative AI, agents, RAG and MLOps capabilities.
itsmsoftware.plContracts, support and documentation in English or Polish.
Years of delivery in public sector and finance.
Mint Service Desk and connectors built in-house.
30 minutes with our AI architect. No sales deck — we ask about your use case, data and infrastructure, and tell you what's realistic.