Data Power has launched an AI agent designed to help marketplace sellers identify hidden financial losses. The tool automatically scans personal accounts to detect profit leaks and provides managers with actionable instructions. It is specifically targeted at medium and large businesses with a monthly turnover starting from 20 ₽ million.
The system utilizes a hybrid architecture to ensure accuracy. Fixed software code handles calculations and anomaly detection, while AI is employed only at the final stage to interpret mathematical findings into user-friendly language. The AI operates based on 20 strict prompting rules that account for business context and prevent technical errors.
The agent monitors advertising, sales, and inventory management, identifying issues such as campaigns where ad spend exceeds margins. Data is collected via API, anonymized, and stored on secure servers located in Russia. The tool currently reduces the time required for anomaly detection from 8 hours to 15 minutes.
The platform incorporates a human-in-the-loop approach, where financial analysts review the agent's findings. User feedback is then fed back into the system to train the neural network and improve the accuracy of future analyses.
