Case Study: InPost scales AI-driven operations and next-day delivery with Dynatrace

InPost resolves Kubernetes issues in 7 minutes with Dynatrace

InPost, a major European logistics company processing around 1.4 billion parcels annually, faced the challenge of scaling its AI-driven operations to meet its ambitious goal of 98% next-day delivery rates. As it expanded, monitoring the performance and efficiency of its diverse AI systems, including shopping agents and forecasting models, became increasingly complex. To maintain service quality and control costs, InPost needed clear visibility into these AI workloads and turned to the observability platform Dynatrace for a solution.

By extending its use of the Dynatrace platform, including its Grail data lakehouse and MCP capabilities, InPost gained comprehensive AI observability. This provided real-time insights into model performance and user interactions, enabling teams to quickly identify and resolve issues. As a result, InPost optimized token usage for its shopping agent, improved developer productivity, and ensured reliable AI-driven forecasting. The solution delivered significant operational efficiencies, exemplified by a Kubernetes issue that was identified and resolved within seven minutes, preventing any customer impact and supporting InPost's next-day delivery goals.

InPost