Askona Freed Up 130,000 Working Hours: How an AI Agent Automated Finance Department Routine

In a large company, financial operations and analytics involve thousands of hours of manual labor: collecting data from disparate sources, endless reconciliations, creating templated reports, and identifying anomalies. This routine does not create value but consumes the time of qualified specialists who could be engaged in strategic planning and in-depth analysis. The Russian sleep products manufacturer, Askona, faced this challenge and successfully implemented AI agents, freeing up the equivalent of 130,000 working hours and shifting the focus of its financial and analytical departments from data collection to strategic tasks.

In a large company, financial operations and analytics involve thousands of hours of manual labor: collecting data from disparate sources, endless reconciliations, creating templated reports, and identifying anomalies. This routine does not create value but consumes the time of qualified specialists who could be engaged in strategic planning and in-depth analysis. This way of working is not necessary; today, this burden can and should be automated.

How the Finance Department Operated Before AI Agents

Before the implementation of AI agents, Askona's financial and analytical departments operated under a classic scheme. Employees daily engaged in routine information gathering from various corporate systems – ERP, CRM, banking platforms, and internal databases. This process required significant time investment and involved manual extraction, consolidation, and reconciliation of data.

After collection, the data was manually processed, checked for compliance with numerous internal regulations, and only then compiled into reports. This work was not only time-consuming but also prone to human error. The diversity of data sources, the need for constant reconciliation, and manual input increased the risk of errors, which in the financial sector can be very costly for a company.

Highly qualified specialists, hired for analysis and strategic planning, were forced to spend a significant portion of their working hours on mechanical operations instead of focusing on identifying new business opportunities, in-depth analysis of market trends, and optimizing financial flows.

Why Existing Tools Were Insufficient

Askona, like any large company, already had various automation and accounting systems in place. However, these tools were designed to perform specific tasks and lacked the ability for complex interaction, natural language data interpretation, or autonomous operation. They could automate individual steps of a process but not the entire cycle.

For example, an accounting system could collect transaction data, but its analysis, reconciliation with CRM metrics, report generation based on specified criteria, and anomaly detection still required direct human involvement. Existing solutions were focused on performing specific, clearly defined tasks. They lacked the ability for complex interaction, natural language data interpretation, or autonomous operation with dynamically changing requests.

The company needed a solution that could not just execute predefined scripts but act as an intelligent assistant, capable of independently collecting, analyzing, and synthesizing information, reducing the burden on employees and minimizing errors. This ultimately led Askona to the decision to implement AI agents. A tool capable of end-to-end automation, which could "understand" context and take on chains of routine operations, was necessary.

How the AI Agent for Finance Was Designed

AI agents were designed as multifunctional systems capable of performing a range of key tasks in the financial and analytical spheres. The core idea was to create an intelligent "orchestrator" that could interact with various corporate systems, collect necessary data, process it, and present it in a human-friendly format. The agents were required not just to automate, but to exhibit intelligent agency – the ability to act autonomously within defined rules.

The functionality of AI agents included:

• Automated Data Collection and Processing. Agents learned to independently extract information from disparate sources such as ERP, CRM, banking systems, and internal databases, which was previously done manually. They not only collected data but also performed primary validation and cleaning.
• Financial Report Generation. Based on predefined parameters, AI agents generated standard financial reports and analytical summaries, including quarterly and annual reports, income and expense statements, and cash flow statements. Agents could adapt report formats to various internal and external stakeholder requirements.
• Anomaly Detection and Monitoring. Agents continuously monitored financial operations, identifying deviations from norms and potentially suspicious activity. This helped not only prevent errors and fraud but also respond promptly to changes in market conditions.
• Analytics Optimization. AI agents took on tasks related to forecasting financial indicators based on historical data and current trends. They performed rapid analysis of large data arrays to identify hidden patterns and correlations, providing analysts with already processed hypotheses for verification.

A key principle was to create agents that could work autonomously yet remain transparent for human oversight. Employees gained the ability not only to receive ready results but also to verify the agents' logic to ensure data accuracy and conclusions. This maintained control and trust in the new system.

Implementation and Adaptation

The implementation of AI agents at Askona was phased, which allowed for minimizing risks and ensuring smooth employee adaptation. In the first stage, the most routine and predictable processes were automated, such as data collection for standard reports and primary reconciliation. This allowed employees to quickly see the benefits of the new system, confirm its reliability, and overcome initial resistance.

After a successful pilot launch, AI agents were gradually integrated into more complex analytical tasks, such as forecasting and anomaly detection. A critical aspect of the implementation was training employees to work with the new tools. Financial analysts and managers learned to formulate queries for the agents, interpret the results obtained, and use them to make more informed decisions.

This approach helped minimize resistance to change and ensured a smooth transition to new workflows. Employees stopped viewing AI agents as a threat and began to see them as powerful assistants, freeing them from tedious routines and allowing them to focus on more intellectual tasks.

Implementation Results

While Askona does not disclose exact figures on cost savings in monetary terms, the implementation of AI agents brought tangible results in terms of freeing up human resources and increasing efficiency:

The use of AI agents significantly reduced the time required to collect, process, and analyze large volumes of financial information. What used to take hours or even days is now completed in minutes. Automation of routine operations minimized the impact of human error, leading to a substantial reduction in the number of errors in reports and analytical summaries.

Employees in the financial and analytical departments were freed from routine tasks and could focus on more strategic, creative, and intellectual activities, such as in-depth analysis of market trends, development of new financial models, and optimization of business processes. By rapidly processing vast amounts of data and identifying hidden patterns, AI agents allowed Askona to gain a deeper and more accurate understanding of its financial performance and make more informed decisions.

Overall, the implementation of AI agents at Askona led to an increase in the overall efficiency of financial and analytical functions, which directly impacted the company's competitiveness.

How to Implement This in Your Company

Askona's case proves that AI agents can be a powerful tool for optimizing financial and analytical processes in any company with a large volume of routine operations. If you want to replicate this success, start with a few key steps:

• Identify the most routine tasks. Make a list of all repetitive operations in your financial and analytical departments that consume a lot of time from qualified specialists. This is an ideal candidate for the first AI agent.
• Start small. Don't try to automate everything at once. Choose one or two tasks with clear rules and measurable results for a pilot project. This will allow you to quickly see the return and confirm the solution's effectiveness.
• Integrate agents into existing systems. The less employees have to change their habits and learn new interfaces, the faster they will adopt AI agents and actively use them.
• Focus on support, not replacement. AI agents should be assistants that handle routine tasks so that people can focus on decision-making and strategic planning.

If this case sounds like what's happening in your company, our manager can help: he'll analyze your business and niche for free and point out where an AI agent would bring a real result in your case. Message the manager