1. Home
  2. FAQ
Questions people ask

Clear answers before you decide

Short, direct answers grouped by who is asking. Where a figure comes from a client project, we say which one; where something is still on the roadmap, we say that too.

Illustrative portrait of a head of customer service in a calm contact centre
For executives

For executives

What NatusAI does, what you can buy today, and how fast you see a result.

Related pages

What does NatusAI do?

NatusAI connects your organisation’s knowledge into AI your people can verify, on infrastructure you control. A senior team starts from one costly business question, proves the result, and builds on a platform you own, with no lock-in to a single model or cloud. See the client results.

What can we buy today, and what is still on the roadmap?

Today you can engage our team to deliver an AI application on the engine proven in our client projects, using NatusAI Lyra interfaces (Available). NatusAI Cygnus (data foundation) is planned for its first release. NatusAI Orion (application platform), NatusAI Auriga (AI coordination), NatusAI Aquila (document intelligence), NatusAI Draco (operations) and NatusAI Gemini (verified recovery) are on the roadmap. Every product carries a status badge, and we share dates under NDA. See products and status.

What results have you delivered?

Three measured client results in regulated sectors: 92% faster customer response at a European telecommunications operator, one of Europe’s first country-wide AI grid digital twins at a national electricity distribution operator, complex queries from 3 days to under a second at a large European bank.

How quickly will we see a result?

We start with one business question and a measured baseline. A scoped discovery phase delivers a costed plan and ends with a clear go/no-go decision before anything is built. A European telecommunications operator moved standard enquiries from 3–5 days to hours with a 6-hour team training and 30 days of post-launch support.

What does it cost?

Commercial terms are agreed per engagement. Discovery gives you a costed plan before you commit to a build, so you see the investment and the expected return side by side. Product pricing, licensing and packaging have not been selected. AI funding and billing are also open; no provider, included allowance, invoice policy or unlimited-use promise is stated. Ask us for a scoped proposal.

Does AI replace our people?

No. AI takes the agreed routine work and routes everything else to people. At a European telecommunications operator, AI handled 12 recurring enquiry types, complex cases went to human agents, productivity rose 30%, and the client highlighted a better working life for employees. Read the telecom case.

Risk, compliance & security

Risk, compliance & security

Where it runs, what leaves your environment, and how every AI-assisted decision stays defensible.

Related pages

A protected infrastructure environment, representing control and responsibility (illustrative render)

Can NatusAI run inside our own data centre?

Yes. NatusAI is designed to run on-premises, in a European sovereign cloud or in a hybrid setup. The national grid digital twin we delivered runs on standard IT infrastructure, with no expensive hardware upgrades. You decide which data and models stay where, and we document it before we build. More on deployment and sovereignty.

Does our information leave our environment?

Only if you decide it should. What leaves depends on the data and AI-provider routes configured for your scope, so we make every route, credential, access right and log explicit before go-live. We do not make a blanket data-residency promise: we write your boundary into the architecture.

How do you prevent AI from giving wrong or unsupported answers?

Assistants answer from sources you approve and show them. Consequential actions require a named person’s approval, and complex cases are routed to people. We test quality before rollout and improve it with real cases.

Does AI act without human review?

Not on anything consequential. Our design separates a proposal from the permission to act: AI proposes, and a named person approves. Protected actions need their own authority, and changes pass verification before release. Routine work runs automatically only within boundaries you agree.

Can investigators and auditors see why the AI flagged something?

Yes. Each result comes with the records, relationships and history behind it, so people review the context rather than accept a black-box score. At a large European bank, investigators see the transactions, counterparties and history behind every alert.

Does NatusAI help with the EU AI Act, GDPR, DORA and NIS2?

We support your obligations with human oversight, logging, source transparency, on-premises or EU deployment, and documented exit and recovery plans. We do not certify your systems or replace your legal assessment. Our governance pack maps these features to each framework. See governance and compliance.

Is NatusAI certified to ISO 27001 or SOC 2?

We do not claim any certification on this site. Ask us for our current certification status and our security documentation; we share them before any procurement step.

Will we be locked in to NatusAI or to one AI provider?

No. You own your data and the applications we build for you. NatusAI is model-agnostic: we use hosted, EU-hosted or local models according to your policy. An exit plan is part of every engagement from the start.

Does a backup mean recovery is proven?

No. A completed backup and a verified restore are different things. Restore drills you can show to an auditor are what count. Verified recovery is the role of NatusAI Gemini (Roadmap); until then, recovery testing is agreed and evidenced per engagement.

For technical teams

For technical teams

Models, integration, architecture and what the first platform release does and does not include.

Related pages

Which AI models do you use?

The ones your policy allows: hosted, EU-hosted or running locally. Past projects have used leading commercial models. Where data must not leave your environment, we use models that run locally. Model choice is recorded in the architecture and can be changed without rebuilding the application.

Do we have to replace our CRM, document management or ERP?

No. We build into the systems you already run. At a European telecommunications operator, the solution was integrated with the existing CRM and ticketing platform with no workflow disruption. See how the platform connects to your systems.

Do we need new hardware?

Usually not. A national electricity distribution operator runs a country-wide grid digital twin on standard IT infrastructure, and no expensive hardware upgrades were needed. Read the energy case.

What is NatusAI Orion?

NatusAI Orion is the vision for a connected application platform where people describe what they need and AI helps with application development and operations. It is a product vision (Roadmap), not a claim that the full platform is currently available. Read the Orion vision.

What is NatusAI Cygnus?

NatusAI Cygnus is designed as a programmable temporal graph database and backend that connects entities, relationships, time series, geospatial data and retrieval. Worldwide digital twins are its first demanding domain, not its only intended use. Explore NatusAI Cygnus.

What are the seven NatusAI products?

Seven roles on one platform: NatusAI Orion (application platform), NatusAI Cygnus (connected data foundation), NatusAI Lyra (interfaces), NatusAI Auriga (AI coordination), NatusAI Aquila (document intelligence), NatusAI Draco (operations) and NatusAI Gemini (verified recovery). They describe responsibilities, not seven separate subscriptions. Only NatusAI Lyra is available today. See all products.

What will the first release of NatusAI Cygnus include?

NatusAI Cygnus is the planned data foundation: it connects entities, relationships, history, location and search in one place. Version 1 is planned with embedded, development and single-server production profiles. Clustering, multi-node failover and distributed execution are later scope. Explore NatusAI Cygnus.

Does “worldwide” mean multiple servers?

No. The planned 1.0 production profile is single-server. It can represent globally distributed entities within that node’s capacity, but cluster, multi-node failover and distributed execution are later scope.

What deployment profiles are planned?

The planned NatusAI Cygnus profiles are embedded, development and single-server production. Cluster and isolated-workload profiles are not supported in v1. A hosting environment a customer prefers does not by itself establish that a product profile is supported. See the planned deployment profiles.

Are the Cygnus performance figures proven?

Not yet. Workload figures for Cygnus are design targets and release gates, not measured results. They must be benchmarked on a stated workload, hardware and topology before any result is claimed. The results we do claim come from delivered client projects.

Can it process every document or media format?

No universal promise. The first release does not include a universal parser, OCR or media-decoding suite. Exact formats and processing providers are qualified per scope, against your actual documents.

Working with NatusAI

Working with NatusAI

The team, the engagement, adoption and support.

Related pages

Who will we work with?

A small senior team: a lead consultant who owns the business outcome, a solution architect, and engineers who build inside your environment, plus an enablement lead for training.

How does an engagement work?

Three decisions, four phases. Discovery (analyse) frames one costly question and checks the evidence. A pilot (design, build and integrate) proves the result inside your environment against agreed KPIs. Scale (adopt and improve) trains people team by team and reuses the same foundation. Each stage ends with a decision you control. How we work.

Who owns what you build?

You own your data, the models you choose and the applications we build for you. They run in your environment, and we plan and document the exit at the start of every engagement, not at the end.

How do you get thousands of people to use it?

Specialised assistants behind one familiar interface, practical training and a department-by-department rollout.

What support is available after go-live?

Support levels, response times, access methods and update responsibilities are agreed for each engagement and deployment, and written into the scope.

How do we start?

Write to contact@natusai.com or choose a conversation: an executive briefing, a use-case discussion, a technical deep-dive or a partnership. Bring the business question that costs you most; no preparation is needed.

Build your next chapter

Still deciding where to begin?

Bring the question your organisation is already asking. We can help define a sensible scope and the next decision.

Or write to contact@natusai.com.

Financial-crime investigators reviewing transaction tables and a relationship diagram together (illustrative image)