Why Positive Chose to Focus on Artificial Intelligence

Positive chose to focus on artificial intelligence because the way companies use knowledge, data, software and employees' time is changing. We do not regard AI as an isolated technology experiment. It may help connect and simplify parts of everyday operations, but only when the organisation knows which problem it is solving and who owns the result.

This article restores the theme of an earlier Positive blog post credited to Miljan Radanovic. The original author's perspective described why the company directed part of its attention and development towards AI, referring to business consulting, digital transformation and the earlier PAM software context. The core reasoning is retained here, while references to today's Positive ecosystem are aligned with current public service descriptions.

Why did Positive make AI a strategic focus?

The earlier article argued that artificial intelligence was not simply the next feature to add to a software package. It could change the way organisations find knowledge, prepare decisions and carry out repetitive work.

Making AI a strategic focus is not the same as believing every company must purchase an AI tool immediately. It means investing in understanding the technology, developing skills, designing useful services and learning how to assess when AI adds value and when it does not.

Companies can adopt several impressive products yet continue working with fragmented information and slow processes. A more lasting advantage is the ability to identify useful problems and design a responsible way to solve them.

Three choices described in the original article

The historical Positive essay presented three possible responses: begin exploring early, wait until the wider market adopts the technology or ignore the shift. These alternatives expressed the author's perspective at the time. They should not be treated as a scientific framework or proof that every early adopter wins.

Early exploration can help teams learn what systems do, where they fail and which information they need. Yet experiments without a clear objective can waste resources. Waiting may avoid certain mistakes but also delay learning. Ignoring a technology that materially affects customers and work can create risks of its own.

Our practical conclusion today is to avoid both uncritical haste and indefinite delay. The right time to invest depends on a specific business case, the readiness of data and the ability to manage risks.

Artificial intelligence is not a substitute for digital transformation

Digital transformation involves workflows, responsibilities, organisational practices, data and technology. AI can play a role in it, but cannot automatically repair conflicting records, unclear procedures or software that does not exchange information.

If sales requests live in chat messages, service teams keep separate spreadsheets and management has no shared operational view, adding an AI assistant will not by itself create a functioning system. Someone must first decide which record is authoritative, who maintains it and how changes move between teams.

Positive's current presentation describes a process beginning with business problems and priorities, then combining AI, business software and dependable IT foundations. This is why our AI focus is part of a broader transformation strategy rather than an isolated marketing promise.

Where AI can help in everyday business operations

Finding knowledge. Employees may spend time looking through documents or asking colleagues for information. An assistant connected to approved, accessible sources can support faster discovery of relevant guidance.

Preparing communication. AI can summarise a request, suggest a response or classify an enquiry. People remain responsible for accuracy, promises and the final communication.

Reducing repetitive administration. When teams repeatedly copy information between applications, automation may help. Sometimes straightforward software integration is sufficient; AI is most relevant when part of the task requires interpreting unstructured text or content.

Working with reports. An assistant can help summarise reliable figures and formulate useful questions. It cannot prove what caused a change merely by writing a confident explanation.

Customer support. A chatbot with maintained service information can handle common requests and prepare a useful handoff to an employee. Without an escalation process, a continuously available chatbot may simply repeat a problem rather than solve it.

All these applications require clear responsibilities, quality checks and meaningful measurements.

Why Positive does not rely on one AI product

AI operates within a business environment. It needs reliable information, workflows with owners and infrastructure that supports daily operations. One organisation may need an employee assistant; another needs better CRM processes, task coordination or more resilient IT first.

That is why Positive currently describes a connected ecosystem. Cybercompany focuses on AI strategy, assistants, chatbots and automation. ONE provides a modular business platform for connected customers, tasks, projects, documents and requests. CoreTech addresses infrastructure, IT support, continuity and cybersecurity.

This does not mean each customer needs all three areas. The point is to select the elements justified by actual business needs and establish clear ownership over the plan and its results.

Explore the Positive ecosystem for the wider relationship between these services.

Cybercompany's role within the ecosystem

Cybercompany represents the specialised AI area of the Positive ecosystem. Its current public offering covers assessing business opportunities, developing AI strategy, building assistants and chatbots and implementing controlled automation.

Instead of beginning with the question “Which model should we buy?”, the work starts by understanding where time is lost, which documents and systems exist and what change would be worth an investment. A first use case can then be selected, measured and improved before wider deployment.

This helps organisations move beyond isolated demonstrations. AI becomes a plan that accounts for users, information, technical systems and controls. See Cybercompany AI solutions for the current offering.

Processes and data matter more than the model name

AI systems can generate persuasive language even when the input data is incomplete or wrong. If two departments maintain incompatible versions of the same contract, an assistant without source rules may rely on the wrong document. Without ownership, it may also be difficult to correct the underlying record.

Serious deployments therefore need at least four foundations: a defined workflow, reliable sources, suitable access permissions and someone accountable for important decisions. These requirements support reproducibility rather than adding bureaucracy for its own sake.

Sometimes the right first step is to organise documentation or integrate existing applications. Introducing generative AI after these improvements can be more effective than attempting to automate a disorderly process.

Employees remain essential to AI adoption

The original Positive article highlighted employee training and the acceptance of change. Both remain essential. A tool that staff do not understand, cannot trust or must constantly correct may end up being underused.

Training should include practical examples from everyday work, data-handling rules, methods for checking AI output and a way to report errors. Leaders should explain what is changing, which decisions remain human responsibilities and how results will be evaluated.

Automation should not be described solely as replacing people. In many processes the more useful goal is reducing copying, searching, waiting and avoidable repetitive steps, giving employees more time for judgment and customer relationships.

AI strategy must include risk management

AI carries risks. Confidential information can be entered into an unapproved tool. Models can produce mistakes. Automated actions can be executed without appropriate controls. Employees can rely on a plausible answer that is unsupported by evidence.

The NIST AI Risk Management Framework encourages organisations to govern, map, measure and manage AI risks. In practical terms, a business should assign owners, define acceptable uses, test quality and monitor outcomes after deployment.

Controls need to reflect the task. Drafting a marketing headline and changing a financial record do not carry the same risk. Consequential or sensitive work requires extra checks, constrained permissions and suitable human oversight.

How should the first AI project be selected?

Begin by mapping a real workflow: who receives a request, where information is found, how many steps processing requires and which delays repeat. Once those details are visible, it becomes easier to assess whether AI could improve the process.

Next, examine readiness: documents, data quality, access permissions, integrations and the people who will use the system. Potential use cases can then be ranked by business value, feasibility, risk and adoption requirements.

For the selected case, define a pilot with a baseline and meaningful success criteria. If the pilot demonstrates value, plan for ongoing ownership, training and gradual expansion. If it does not, adjust the process or stop rather than presenting a demonstration as a successful deployment.

This approach reflects the current logic of Cybercompany's AI strategy offering, which connects business problems, readiness and accountability.

How can a company measure useful AI outcomes?

Time to a completed task is more relevant than how quickly a model generates text. Saving five minutes on an initial draft does not create a net gain if employees spend ten minutes repairing errors.

Output quality can be measured through accuracy, completeness, review cycles and customer experience. Costs include subscriptions, integration, maintenance, training and monitoring. Adoption should show whether employees use the system because it helps them, rather than merely because they have been told to.

For more complex systems, risk indicators also matter, such as unauthorised access or failed escalation. Without a baseline, it is difficult to attribute improvements to a specific tool.

What does Positive's original AI decision mean today?

The historical post argued that the company should understand and invest in emerging technology rather than wait until change could no longer be ignored. It referred to earlier software products, expectations and a specific development context. Those historical statements should not be read as guarantees, independently verified performance figures or descriptions of every current product capability.

Today's explanation is more operational: use AI when there is a genuine need, sufficiently prepared information, a responsible team and a method of measuring success. Technology matters, but its place within daily work matters more.

The best outcome is not the company with the most AI subscriptions. It is the organisation that finds reliable information faster, spends less effort on avoidable manual tasks and retains control over important decisions.

Conclusion: focusing on AI means improving the way work gets done

Positive invested attention in AI because it offers new ways to connect organisational knowledge, business processes and technology. The reason is not fashion, nor the promise that a chatbot will solve every problem.

When strategy, people, data, business applications and IT foundations are aligned, AI can become a useful part of digital transformation. When those foundations are missing, addressing them may be the most productive first step.

To evaluate realistic AI opportunities, visit Cybercompany solutions or contact the Positive team. The journey starts not with purchasing a tool but with choosing a business problem worth solving.