Implementation of the Zendesk ecosystem for optimising processes and inter-team communication within the technical support department, as well as the integration of an artificial intelligence agent in a product-oriented IT company

This thesis presents a comprehensive implementation of the Zendesk ecosystem aimed at optimising technical support processes, strengthening inter-team communication, and introducing intelligent automation within a product-oriented IT company. Conducted at Proxima Research International, the project involved the migration from a legacy ticketing system to Zendesk, the development of a unified support workflow, and the integration of complementary tools, including Jira for incident escalation and the internal CRM for synchronising customer data. A central component of the work was the deployment of an AI-driven chatbot, designed to automate the processing of recurring requests, leverage the knowledge base through structured conversation flows, and provide multilingual, 24/7 user support. The implemented solution standardised support procedures, improved SLA compliance, and enhanced visibility across departments through shared dashboards and automated reporting. KPI-driven evaluation showed measurable improvements: reduced ticket resolution times, increased first-response consistency, and an automation rate of up to 18–21% of monthly requests. Knowledge base optimisation further contributed to reduced agent workload and faster incident diagnosis. The combined ecosystem enabled seamless data exchange, transparent prioritisation, and a more efficient escalation pipeline between Support, Product, and Customer Success teams. The findings confirm that integrating Zendesk with AI-based automation significantly increases operational efficiency, customer satisfaction, and process scalability. This work provides a practical reference model for IT organisations aiming to modernise technical support through intelligent service platforms, workflow standardisation, and cross-system integration.