What SenAi is and what it is for

SenAi is a system under development for collaboration between people and artificial intelligence agents, intended to carry out knowledge-based tasks and processes. It combines AI models’ reasoning with persistent data, recorded operating rules, execution tools and user control.

Its purpose is to take a task from an identified need to a verified result: establish what needs to be done, gather the necessary information, prepare a course of action, perform authorised operations and present the outcome together with an account of the work.

The fundamental idea behind SenAi is to reverse the traditional approach to implementing information systems. Rather than people learning to operate a system and adapting their processes to it, SenAi learns from people: their ways of working, knowledge, experience and the rules governing their organisation.

A system that learns about the enterprise

SenAi is installed in an initial form, without knowledge of the enterprise, its organisational structure, documents, products, customers or business processes.

This knowledge is acquired gradually during everyday use.

Employees perform their regular duties in collaboration with AI agents. Through this collaboration, SenAi learns about tasks, documents, dependencies, procedures, decision rules and ways of solving problems.

The system does more than store information. Its role is to recognise its meaning, relationships and potential for reuse. Acquired knowledge is recorded, verified and made available for subsequent tasks.

In this way, SenAi gradually becomes an electronic representation of the enterprise — its knowledge, processes, experience and ways of operating.

This is neither a static database nor a model of the organisation prepared just once. It is a structure that develops alongside the enterprise’s activities, incorporating new information, organisational changes and employees’ experience.

SenAi is therefore not a ready-made electronic model of an enterprise. It is a mechanism that gradually builds such a model through everyday collaboration with people.

Processes developed during work

One of SenAi’s central design principles is the ability to create business process scenarios while those processes are actually being carried out.

In a traditional approach, implementing a process requires prior analysis, documentation, solution design, software development and testing.

In SenAi, the starting point is the user’s actual work.

An employee carries out a task with AI agents, explains decisions, identifies required information and approves subsequent actions. The system analyses the workflow and recognises recurring activities, dependencies and operating rules.

It can use this information to propose a process scenario which, once verified and approved, can be reused for similar tasks.

A process scenario need not be created before work begins. It can emerge while the work is being performed, based on employees’ actual experience.

Across successive executions, SenAi can recognise deviations, identify opportunities for improvement and propose changes to existing scenarios.

Every employee can participate in developing the system in this way. An accounting employee contributes knowledge of accounting processes, a salesperson of sales processes, and a logistics specialist of logistics processes. This does not require programming skills or knowledge of SenAi’s internal architecture.

Radically reducing process implementation and maintenance costs

Implementing SenAi does not require taking employees away from their daily duties for lengthy interviews, process descriptions and analytical documentation, followed by programmers translating that documentation into code, testing it and correcting the software.

SenAi learns processes while they are actually being carried out. Employees perform their regular tasks while the system, working with them, gathers knowledge about workflows, decisions, documents and applicable rules. It uses this knowledge to prepare scenarios that can be reused after approval.

This approach is intended to radically reduce the time and cost of analysing, implementing and subsequently maintaining business processes. It also reduces the enterprise’s dependence on large teams of analysts, programmers and testers.

Crucially, SenAi is intended to recognise process changes during everyday work.

If an employee begins carrying out a task differently from the existing scenario, the system can detect the deviation, establish its cause and propose an update to the process.

Once the change is approved, the new scenario becomes available to other authorised SenAi users. There is no need to organise lengthy workshops again, recreate documentation from scratch or commission code changes, provided the change is within the capabilities of the existing execution tools.

As a result, the enterprise and its electronic representation in SenAi can develop in parallel. Changes in the real organisation can be continuously recognised, verified and reflected in the system.

In traditional systems, a business process change often requires a software change. In SenAi, it may primarily require updating knowledge and an operating scenario without modifying the system’s code.

This represents a move away from software that must periodically be adapted to a changing enterprise towards a solution that continually learns how the enterprise operates and can adapt its scenarios accordingly.

The economic benefits are therefore not limited to initial implementation. They also concern long-term maintenance, development and adaptation to changing business needs.

Collaboration between people and AI agents

SenAi is intended to serve both individual users and entire organisations.

One example is preparing a quotation using a customer enquiry, product documentation and company rules. In such a process, the system should identify requirements, flag missing data, use the appropriate template, prepare the document and obtain required approvals.

A similar approach can support analysis, documentation, project knowledge management and recurring tasks.

Work is divided among collaborating agents. ALEX interprets tasks and coordinates the work, GREG assesses the matter submitted for review, and MAJA carries out the agreed task.

Their actions follow recorded rules. The user retains control over the objective, scope and decisions requiring their consent.

The user does not need to know SenAi’s internal architecture or design algorithms. Their primary contribution is knowing how to do their work and providing the knowledge the system needs to support it.

Employees should not have to adapt the way they perform their duties to SenAi. SenAi should learn how those duties are performed and learn to support them.

Shared enterprise knowledge

Knowledge gathered by SenAi need not remain tied exclusively to the employee who contributed it.

Employees’ experience, problem-solving methods and knowledge of processes and operating rules can become lasting organisational assets.

Knowledge need not disappear when an employee leaves, changes position or the enterprise is reorganised.

Information acquired through tasks in one department can also be used elsewhere in the organisation, subject to appropriate permissions.

In this way, SenAi can gradually recognise cross-departmental processes, identify dependencies between departments’ activities and reconstruct workflows spanning the entire enterprise.

An important element of this approach is distinguishing information from binding rules. Not every statement contributed by an employee should automatically become an operating principle.

The system should account for the source of knowledge, its currency, its scope of applicability and potential conflicts with previously recorded arrangements.

Knowledge continuity and oversight

A central principle of SenAi is continuity of work. Data, agreements, scenarios and results should remain available beyond an individual conversation with a model.

Subsequent tasks can therefore use previously accumulated knowledge rather than reconstructing the entire context each time.

Knowledge acquired through one employee’s work can be used by others in the organisation, subject to appropriate permissions.

Another principle is traceability of actions. The system should distinguish an intended operation from its actual outcome, retain information about errors and stops, and make it possible to establish which data and rules underpinned a result.

This supports both oversight of task execution and analysis of errors and improvements to processes.

Improvement through experience

SenAi provides for improving its ways of working through experience gained during task execution.

After a task is completed, the system can analyse its course, recognise a change or deviation, establish its cause and prepare an improvement proposal.

Implementing such an improvement is subject to appropriate approval and verification rules.

SenAi is therefore intended not only to gather knowledge of how an enterprise operates, but also to use experience to propose increasingly effective ways of carrying out its processes.

The system can also identify recurring problems, unnecessary activities, inconsistencies in working practices and opportunities to improve cooperation between employees and departments.

Every completed task can therefore meet a specific business need while also providing new knowledge for improving the organisation as a whole.

Relationship to AToS

SenAi’s relationship to the Axiomatic Theory of Self-Consciousness (AToS) primarily concerns designing a system capable of recognising changes in its own operational state, identifying their origins and analysing the consequences of its own actions.

This is a direction for SenAi’s construction and research. Adopting this architecture does not in itself establish that the self-awareness criterion has been met.

Summary

SenAi is intended to be a system that learns about an enterprise through its everyday activities.

It starts without knowledge of the organisation and gradually learns its structure, employees, processes, documents, rules and experience.

This knowledge is not merely archived. It is intended to support subsequent tasks, the creation of scenarios for recurring processes and proposals for improvement.

The enterprise need not first describe all its processes in detail and then commission their implementation in software. Scenarios can emerge during everyday work, and changes can be recognised and implemented without involving software development teams on every occasion.

This approach is intended to significantly reduce the costs of implementing, maintaining and developing systems supporting business operations.

Employees should not have to learn SenAi. SenAi should learn from employees.

The enterprise should not have to adapt to an information system. The system should learn about the enterprise and develop alongside it.

This description presents SenAi’s purpose and design principles. Confirming the scope of implemented capabilities and the scale of economic benefits requires separate verification of the implementation and deployment results.