01 — Vision
A world in which every person works in a duet with artificial intelligence and devotes their time to where they are irreplaceable.
02 — Mission
We build organisations that work with artificial intelligence in a managed way — secure, lawful, economically justified and maintained over time.
Each of these four terms is a condition met by every AI deployment we deliver.
010b001
Secure
Data stays under the organisation’s control. We agree in advance what the system does on its own and what a person approves.
020b010
Lawful
System register, risk classification, oversight and transparency. Compliance is part of the project from day one.
030b011
Economically justified
Every deployment has a calculated baseline and an expected result. We deliver solutions that pay back within an agreed time.
040b100
Maintained over time
After launch we maintain solutions, follow developments in technology and regularly adapt them to the client’s needs.
03 — Duet
Why we are called Duonaut
Nauta means sailor, duo means two. Duonaut is two sailors on one deck, heading for waters no one has charted yet. One of them is a human. The other — artificial intelligence.
The name expresses the principle on which the whole company rests and is the measure of every decision we make: a good solution strengthens people and works together with them.
Human · analogue signal
Judgement and responsibility
Context, relationships with other people, decisions about direction
Artificial intelligence · digital signal
Repeatability and scale
Patience, consistency and the same quality of work at any scale
The coming decade will belong to duets
The advantage will go to organisations that combine human judgement with the capabilities of artificial intelligence so that each strengthens the other.
We believe there is a natural division of labour between people and machines. The machine is unmatched in repeatability, scale, patience and consistency. The human is irreplaceable in judgement, empathy, understanding context, sensing other people, creativity and taking responsibility. These competencies remain the domain of people, whatever the computing power.
Today, people hired for their experience and judgement spend a large part of their time on tasks that could be entrusted to a machine. This is a question of allocating the most valuable resource an organisation has.
Why now
Artificial intelligence is the first technology that makes it possible to restore the right division of labour. Earlier waves of automation took over what could be written down as instructions. Artificial intelligence reaches further — into work that requires an understanding of context, which takes up most of people’s time in an organisation.
04 — Governance
Governance is a condition for speed.
An organisation that knows where its data goes, who owns each solution and what happens when a vendor changes scales artificial intelligence with the full support of its security, legal and audit functions.
For us, security goes hand in hand with innovation. We work with the latest tools and build at the highest available technical level — with full knowledge of what we build, why, on whose data and who will maintain it.
05 — How we work
We deploy artificial intelligence the way critical services are managed.
For well over a decade we made sure that systems thousands of people depend on ran predictably. We bring the same discipline to artificial intelligence deployments.
010b001
A product with a lifecycle
A model-based solution evolves throughout its operation. It receives an owner on the client side, a change backlog, versioning, usage metrics and a retirement date agreed in advance.
020b010
A service in the organisation’s catalogue
Every solution has a definition, support level, incident reporting path and escalation procedure in the service catalogue. For every incident it is clear who handles it, how quickly and who makes the decisions.
030b011
Managed change
A new model, a revised system prompt, a new data source — these are changes to a live system. They go through change management, with testing and the option to roll back.
040b100
Improvement as a rhythm
A regular review of what works, what needs improvement and what has changed in technology is a fixed point in the calendar.
06 — How we measure
A number before, a number after
Every promise we make is assigned a unit of analysis: a specific process, role or team. This makes each one measurable and verifiable with a concrete number.
What we measure
- process cycle time
- human effort per unit of output
- quality measured by errors and escalations
- throughput with the same headcount
- time to decision
- unit cost of operation
Business case
Before building, we estimate the cost of development, model operation and maintenance, and compare them with the calculated value recovered. The measure of success is a real return on the deployment.
Every project starts with a baseline measurement.
07 — Principles of cooperation
Clear principles from day one.
We work on projects with a defined goal and a measurable result.
We build software for the adoption of artificial intelligence.
We specialise in solutions based on artificial intelligence; classic, deterministic process automation (RPA) lies outside our scope.
We are paid only by our clients, so in choosing technology we are guided solely by their interests.
We base every project on a measured baseline.
We deliver every solution with governance, documentation and an owner on the client side.
We deploy in agreement with the client’s IT department. All tools and models operate with the organisation’s full knowledge and under its control.
We take on projects in which artificial intelligence solves a specific business problem.
We assess data quality before starting work and build on data that is fit for the purpose.
08 — Our commitments
Seven things we promise.
Clients are entitled to hold us to them.
- 001
We build managed artificial intelligence
Every solution has an owner, documentation, a described data flow and planned maintenance. This is the standard for every deployment we deliver.
- 010
Client data stays with the client
We work with models running in the client’s infrastructure, in an agreed processing region. Any change to this scope requires the client’s informed, written consent.
- 011
We prove it with numbers
Every deployment starts with a baseline measurement and ends with a measurement after the change. We deliver solutions that pay back within an agreed, reasonable time — in line with the analysis carried out before the start.
- 100
We advise on technology independently of vendors
We are paid only by our clients, so we recommend the technology that best meets their needs.
- 101
We recommend AI where it is the right answer
For every problem we identify the most suitable path: artificial intelligence, classic automation, process improvement or a management decision.
- 110
We stay with the organisation for the long term
We maintain solutions and the team’s knowledge, and return at a steady rhythm with a review of what has changed in technology.
- 111
We build the client team’s independence
We document solutions so that the client’s team can take them over in full and develop them independently.