One of the key challenges during an economic slowdown is the need to increase efficiency and competitiveness while simultaneously reducing resources. Process digitalization, meaning the use of modern technologies to automate and streamline daily operations, can bring significant benefits. A major area of potential lies in the increasingly widespread application of tools based on large language models.
Economic slowdown as a catalyst for change
In the face of an economic slowdown, company management can apply various strategies. These might include actions such as “waiting it out,” reducing employment or halting investments. However, there are also businesses that use this challenging time to review and optimize existing processes and look for growth opportunities. A review of existing processes helps identify areas that are inefficient and generate disproportionately high costs.
Investing in technology during difficult times might seem risky, but it is precisely then that optimization and automation of processes can bring the greatest benefits. Companies that choose digitalization have a better chance of emerging from the crisis stronger and more competitive.
How to start the process of digitalization?
Digitalizing processes doesn’t have to involve a revolution. From our experience, the best results are achieved through a gradual process that begins with analyzing current operations and identifying areas that need optimization. It’s worth considering which processes generate the highest costs or take the most time, and starting with automating those.
Key steps are:
- Business process analysis – Understanding which processes can be automated and where digitalization will bring the greatest benefits.
- Choosing the right tools– Dedicated solutions based on neural networks or machine learning can help automate repetitive processes and forecast estimates, which can significantly improve company operations.
- Team training– To ensure the new solution provides the expected benefits, it’s essential to involve the team in its development and then prepare them to work with the new tool, allowing them to fully harness its potential.
How to analyze processes in search of innovation opportunities
Before choosing the tools, it’s essential to create a detailed process map to understand how information and resources flow within the organization. This can be done using flowcharts, which help identify different stages and decision points. If the processes are already mapped out, it’s worth reviewing whether they are still current, as certain elements may have changed in practice and could function differently than, for example, five years ago.
Next, each stage of the process should be examined for its efficiency. This can be done by collecting data on factors such as execution time, handling complaints, logistics costs, inventory management, or customer satisfaction. These data should be compared with industry best practices. Based on this, it will be possible to identify where delays or inefficiencies occur. Bottlenecks can limit the efficiency of the entire process, presenting an opportunity for innovation.
It’s also worth looking at the actions of competitors and current market trends. Technological innovations such as automation, artificial intelligence, or big data can offer significant benefits.
Based on the insights gathered during the analysis, you can precisely define the goals you want to achieve. Not all processes can be automated immediately. Therefore, it’s important to determine which processes have the greatest impact on operational efficiency or generate the highest costs. The priority should be those that will offer the fastest return on investment in digitalization.
Introducing new software into a company
Armed with full knowledge of the processes and areas we want to improve, we can begin the search for the best solution. Choosing the right digitalization tools is crucial to the success of the entire process. The system must be aligned with our goals, while considering the existing solutions already operating within the organization. It’s important that these systems are compatible.
There are many software solutions available on the market that can help in various areas, including:
- ERP systems (Enterprise Resource Planning) – Integrated systems for managing a company’s resources (production, finance, logistics, sales).
- CRM (Customer Relationship Management) – Tools for managing customer relationships, which automate sales and marketing processes.
- Office automation – Tools for automating repetitive tasks, such as invoice processing, reporting, or resource booking.
- Project management systems – Platforms like Trello, Jira, or Asana, which help coordinate tasks and teams.
If your company has unique needs that are not fully met by off-the-shelf software, developing a custom system might be the best solution. Having your own system gives the company full control over its functions, development, and security. It can be adapted to the company’s changing needs, with the ability to add new features, integrate it with other tools, or expand its functionality as the business grows.
While building a custom system involves high initial costs, it can bring long-term savings. A proprietary system avoids licensing fees that are often associated with SaaS (Software as a Service) or the purchase of licenses. Additionally, the company gains independence from external providers, reducing the risk of future costs.
Having a custom system can also provide a significant competitive advantage, especially if it enables the implementation of unique solutions or process optimizations that competitors cannot easily replicate.
Tools utilizing artificial intelligence
Tools utilizing Large Language Models (LLMs) and solutions based on them, such as ‘virtual assistants,’ have rapidly gained widespread use in corporate environments. The key difference between them and previously used chatbots is that the former have a much larger knowledge base and are capable of responding to user queries much more eloquently. Moreover, they can “understand” the context and purpose of the conversation. AI-based technology, therefore, significantly speeds up routine tasks, improving many processes.
There is substantial potential for companies seeking optimization in the development of chatbots and virtual assistants. AI-powered tools are widely used in customer service, allowing companies to automate responses to frequently asked questions, handle inquiries, and provide technical support, reducing the workload of support teams. This enables companies to cut labor costs while ensuring a faster response to customer needs, which is crucial during challenging times.
AI-based programs also allow for analyzing supply chain data, identifying potential bottlenecks, optimizing delivery routes, and managing vehicle fleets. This enables companies to respond faster to disruptions, plan deliveries with greater precision, and minimize transportation costs.
Implementing AI-driven tools can also help with recruitment and employee management by automating processes such as CV screening, evaluating candidates’ competencies, planning work schedules, and analyzing team performance. Automating these processes enables companies to operate more efficiently with fewer resources.
Implementation example
Farm Frites Poland is a leading producer of potato products, supplying its goods to, among others, one of the largest fast-food chains in the world. The company has been operating for over 30 years, with its headquarters located in Lębork. The modern machinery park consists of three production lines. It employs over 270 people, and the quantity of potatoes used for production exceeds 250,000 tons annually.
The problem our client faced was estimating production yield based on the raw material’s parameters—essentially, how much of the final product can be obtained from a given amount of raw material. Simply put, how many fries can be made from a potato. It turns out this is not an easy question to answer, explains Waldemar Korłub, CTO of Inero Software.
With a well-defined set of parameters, we could move on to model research. We tested several different machine learning approaches, including regression models, decision trees, and neural networks. In the end, the best results were achieved with neural networks, with a mean error of around 1.2% for the validation data.
The model was implemented in the client’s infrastructure. Thanks to this solution, the forecast error was minimized to just 3%. This means that accurate estimates can now be made at the stage of contracting raw material supplies.
Preliminary model research was conducted on a subset of data, and the model that performed best was then implemented in our client’s infrastructure. We conducted full model training using the complete set of production data from historical production runs. The final model was deployed to production, achieving a mean squared error below 3% for real-world production runs, emphasizes Waldemar Korłub
Summary
An economic slowdown does not have to mean stagnation—it can be a catalyst for optimization and increased efficiency through process digitalization. Investing in modern technologies enables companies not only to survive challenging times but also to strengthen their market position. As a result, organizations emerge from the crisis stronger, more competitive, and better prepared for future challenges.
