Alexey Khegay shared his experience in implementing AI in property management at the annual CREW Central Asia 2026 forum.

Artificial intelligence entered your company long before its official implementation. Employees use personal subscriptions, upload contracts, prepare reports, compare documents and look for solutions to everyday work tasks.

Along with queries sent to external services, corporate data may also be shared, while valuable experience remains in employees' personal accounts. That is why today, business leaders should be asking not “Are we using AI?” but rather “Who is managing its use?”

According to Microsoft and LinkedIn, by 2024, 75% of knowledge workers were already using AI, with 78% of them relying on personal subscriptions to AI tools for work.

TSPM.'s approach: no hype, just practical value

At TSPM., artificial intelligence is viewed as a working tool designed to solve specific problems and deliver measurable results.

Every AI project begins with an analysis of business processes. It is essential to identify where employees lose time, which operations are repetitive, what data is being used and how the impact of implementation can be measured.

Sometimes, this analysis reveals that a particular business process does not need AI at all, but rather clear procedures, a unified reference database or a properly structured data system. This is also a valuable outcome, as it helps avoid spending resources on technology for technology's sake.

When AI implementation is justified, a solution goes through several stages: technical feasibility assessment, prototyping, MVP development and integration into the workflow. Internal systems do not always require lengthy refinement. The priority is to launch a working version sooner, gather employee feedback and improve the product based on real-world experience.

It is also important to remember that technology evolves rapidly, and spending too much time perfecting a tool may not always be worthwhile. By the time it reaches perfection, it may already be outdated.

People are the driving force behind every project

Even the most advanced technology cannot implement itself.

An AI project needs a leader with the authority, credibility and personal commitment to drive it forward. Not a formal coordinator who receives a monthly report, but someone who is involved in the project's details every day.

This person removes organizational barriers, helps the team adapt established processes and addresses resistance to change. They must be prepared to answer difficult questions and understand employees' genuine concerns.

One of the biggest concerns is the fear of losing one's job. There is no point in avoiding this issue. AI will indeed reduce some repetitive tasks and change the responsibilities of many roles. However, responsibility for decision-making, risk assessment and handling non-standard situations will remain with people.

The AI project sponsor's role is to communicate these changes honestly and demonstrate the personal value of the new tool to employees. Without such an ambassador, even a technically successful pilot risks becoming a system that nobody actually uses.

Nearly 20 AI use cases

Today, TSPM. has nearly 20 processes involving artificial intelligence in its portfolio.

These projects are at different stages, ranging from development and prototyping to MVPs and integration into a unified smart management platform. TSPM. deliberately avoids presenting every demonstration as a finished product. Each solution must prove its accuracy, security and value in everyday operations.

This is particularly important for the real estate industry. According to a JLL study, 92% of corporate real estate teams had already launched or planned to launch AI pilots, but only 5% reported achieving most of their intended goals.

AI starts with well-organized data

The quality of an AI solution directly depends on the quality of the information it receives.

Data contains individual facts about a property, premises, tenant or request. Documents verify these facts. Metadata identifies the author, date, version, status and access rights. Relationships connect premises, tenants, contracts, charges, requests and acceptance certificates into a unified context.

Without this structure, AI receives a collection of unrelated files and may rely on outdated or incorrect information, leading to hallucinations.

TSPM. also places particular emphasis on data confidentiality. Before information is shared with an external service, it is minimized or anonymized. Simply removing names is not enough: a person can sometimes be identified through a combination of their job title, address, date or transaction history. That is why original records are stored separately within a secure internal environment.

Plazusha: an AI assistant in Smart Plaza

One practical example of AI implementation is its integration into the Smart Plaza project.

TSPM. connected artificial intelligence to its internal knowledge base and introduced a digital assistant called Plazusha within the app. It helps visitors get answers to questions about a specific shopping mall, its services, rules and available facilities.

Users no longer need to search through different sections of the app. They can simply ask a question in their own words, either by typing or using voice input. Plazusha accesses the knowledge base and generates a clear, easy-to-understand response.

For visitors, this provides a faster and more convenient way to get the information they need. For the management company, it creates an additional feedback channel: recurring questions help identify which information is mostn demand and which processes may need improvement.

Plazusha does not replace employees. It handles some routine inquiries and helps people get the information they need more quickly.

Innovation as part of everyday business

The main challenge of AI implementation is not choosing the right model. What matters much more is preparing the data, establishing clear rules, integrating the solution into business processes and ensuring that employees actually use it.

This is the approach TSPM. continues to develop: testing hypotheses against real business needs, building prototypes, measuring results and scaling solutions that have demonstrated their value.

Innovation begins when artificial intelligence stops being an impressive demonstration and becomes a managed part of the business, preserving corporate knowledge, supporting employees and making services more convenient for visitors.