On-prem
- Dedicated appliance in your server room or data center.
- Full physical control over hardware and data. Air-gapped configurations available.
- For organizations with their own server infrastructure and strict data residency requirements.
PRIVATE AI · ON-PREM · REGULATED & HIGH-TRUST INDUSTRIESPRIVATE AI · ON-PREM
Private AI for documentation operations in regulated and high-trust industries. On-prem, on dedicated hardware.
Where decisions must be justified, knowledge is scattered and data is sensitive, standard AI tools fall short.
Manuals, repair history, technician notes and expert know-how live in different systems. Every ticket starts with a search.
Work instructions and SOPs drift behind product changes. Updating them takes hours nobody has.
NIS2, AI Act, audit requirements, IP sensitivity. For key entities under NIS2, public AI services are off the table because of data sovereignty.
We create a layer that organizes operational knowledge and supports decisions where security, context, and accountability matter.
It's not about generating impressive answers. It's about moving from scattered knowledge to reliable action.
Full data and process isolation within the organization's infrastructure.
Every answer points to its context, document, and information origin.
Human-in-the-loop for high-stakes decisions. AI proposes, a human approves, the system audits.
From question to process closure, in a controlled path.
It's not about generating impressive answers. It's about moving from scattered knowledge to reliable action.
CortexMine deploys in two models: on-prem on dedicated hardware inside your data center, or in a dedicated single-tenant private cloud. On-prem is our default. In both models the environment is yours alone: dedicated compute, isolated network, data never leaves your perimeter.
Both models run the same tiers and the same flat monthly fee. Tiers, not tokens. Pricing is shared after scoping.
CortexMine connects to your documents where they already live. Changes sync incrementally and the system keeps no copy of source files outside your perimeter. Sign in with your company SSO, no separate accounts.
Document repositories and object storage
Incremental change sync
No copy of source data outside the perimeter
Sign in with your company SSO
Architecture facts, not marketing. No claimed certifications we don't hold.
On-prem on dedicated hardware, or a dedicated single-tenant private cloud under your governance.
Every answer cites the documents it came from.
Every question and answer is recorded and traceable.
Access follows your roles and permissions.
AI proposes, your people approve. High-stakes decisions stay human.
Deployment, updates and hardware refresh handled by CortexMine. No internal AI team required.
Permissions are not a filter placed on a finished answer. A person's role and group decide which documents can enter the answer at all. The same prompt from two people returns results from different sources when their access differs. Control is enforced on every request, at the data layer.
Each Assistant is a configured role on top of our Digital Worker capacity. We start with four areas where CortexMine delivers the strongest operational impact. Audit runs first with our anchor client; the four areas below are built with partners in the Co-Development program.
First line of technical support grounded in failure history, service documentation and OEM manuals. Shortens diagnosis time and unloads senior engineers.
For service managers and Head of Field Operations
Each Assistant is a configured role of a Digital Worker. A Digital Worker is a unit of capacity, not a seat. You pay for work done, not chairs.
Supports the internal audit function on your organization's own documents: file research, evidence retrieval, gap analysis against your procedures. Built for auditability end to end: source-grounded answers, human-in-the-loop, audit-ready log.
In active build with our anchor client, one of the largest Polish manufacturers and a key entity under NIS2.
Audit Assistant works on your own documents. It is not external audit-readiness and does not certify third parties.
Named case study planned.
Explore the Internal Audit use case →
Every answer links to the exact passage in your document, highlighted on the original page. Click a citation and you see the page, the paragraph and the spot the sentence came from. No guessing where the model got it.
Every answer is grounded in your documents and lands in the audit-ready log.
Every part of an answer leaves a trace you can show an auditor.
The question, the sources used, the document version and timestamps saved as a durable record.
Passages unrelated to the question are dropped before the answer is written.
The assistant checks whether it has enough sources and pulls more instead of guessing.
A citation shows the document version from the moment the answer was given, not today's state.
A document is removed from every layer of the system and citations to it are marked as removed.
The AI proposes, your people review and sign off.
One platform, deployed on-prem by default or in a dedicated single-tenant private cloud. Configured for the operational reality of each environment.
Service documentation, instructions, drawing-to-offer and knowledge operations for manufacturers under NIS2.
Explore ManufacturingDocumentation operations for teams that work offline, in demanding and disconnected environments.
Explore Defence & ResilienceSupport for the internal audit function on your organization's own documents.
Explore Internal AuditThe cooperation covers pre-deployment analysis, joint solution design based on feedback, a pilot deployment, and a production deployment.
Ahead of general availability, CortexMine works with a narrow group of industrial partners in a KPI-gated Co-Development program. KPIs are agreed in writing before any work starts.
KPIs agreed in writing before any work starts.
Infrastructure runs at CortexMine's cost during the Co-Development Phase.
The commercial license activates only after the agreed KPIs are met.
Price locked for 3 years from activation.
If KPIs aren't met, we part ways amicably and nothing is charged.
No pilot fee. You pay only after the agreed KPIs are met. If KPIs aren't met, we part ways amicably and nothing is charged.
Commercial availability is rolled out with Co-Development partners first.
After the Co-Development program, we open CortexMine to a wider set of organizations. Four tiers, differentiated by hardware, number of workflows, and concurrent users.
For documentation-heavy operations
Workstation cluster, 5 workflows, up to 60 concurrent users.
Notify me →For multi-site and air-gapped environments in regulated industries
Multi-node configuration, custom workflow scope and concurrency.
Notify me →Flat monthly fee. Hardware and refresh cycle included for on-prem deployments. Tiers, not tokens: no metering, no per-seat pricing. Every tier runs the same frontier-class model in your dedicated environment. Pricing is shared after scoping.
A team of five. Experience in industry, AI, operations and regulation. Backed by an AI advisor: physics PhD from UC Berkeley, research at MIT and Harvard, recipient of the “Outstanding Young Pole” award.
Our team
Built EUR 1.1M+ B2B revenue as CCO at a Polish software house. MSc HCI (Tampere), MBA.
Operations in sensitive-data environments.
Fullstack engineering across RAG, AI and cloud.
Program participant, technology and go-to-market support.
Industry 4.0 accelerator run by Krakow Technology Park.
Program member, access to NVIDIA technology and ecosystem.
Security-focused acceleration program run by Krakow Technology Park. CortexMine is a program participant.
We build CortexMine for organizations that can't afford a simplified approach to AI. We start with manufacturing, and expand into regulated and dual-use environments over time.
On-prem on dedicated hardware is our default model and the one we recommend for regulated environments. We also deploy in a dedicated single-tenant private cloud: the same platform, dedicated compute, an isolated environment. We never run customers on shared infrastructure.
Those are horizontal office tools in a public cloud. CortexMine is a vertical platform for documentation operations, deployed on-prem by default or in a dedicated single-tenant private cloud. Your data never leaves your perimeter.
RAG is a technique, not a product. CortexMine is a platform with configurable Digital Workers, governance, an audit-ready log and productized workflow modules. RAG is one ingredient.
Our anchor client is Grupa Azoty S.A., one of the largest Polish manufacturers and a key entity under NIS2, with the internal audit workflow in active build.
First product version: May 2026. Test deployment: June 2026. Full operational launch and KPI validation: in progress. We publish what is live and what is in build; no inflated claims.
Our clients do not buy cheap AI; they buy AI that fits their compliance duties and customer contracts. On-prem remains our default for regulated environments. A dedicated single-tenant private cloud is available when the organization needs a faster start or has no server room.
Hardware tiers from a single appliance to a cluster, flat monthly fee, hardware and refresh included. No token metering, no per-seat pricing. We share exact pricing after a scoping call.
Estimate workforce capacity and annual savings for your operation.
The cooperation covers pre-deployment analysis, joint solution design based on feedback, a pilot deployment, and a production deployment.
Project implemented under the European Funds for Smart Economy 2021-2027, KPT ScaleUp Booster project.
Grupa Azoty S.A.