Case Study: 2GIS achieves 20% faster incident resolution and reclaims 70 hours per IT specialist monthly with Microsoft Azure Machine Learning

Russian digital maps company kicks technical support into high gear with machine learning

2GIS, a Russian digital maps and guides company serving more than 500 cities and thousands of employees, faced a bottleneck in its internal IT support: two first‑line specialists manually classified over 200 daily tickets across 28 service categories, creating delays and capacity constraints.

Using Azure Machine Learning, 2GIS trained a classifier on 28,000 historical tickets and deployed a solution in about three weeks. Tickets are now auto‑classified in 5–7 seconds (vs. over a minute), saving roughly 70 hours per IT specialist monthly, speeding query resolution by about 20%, and delivering prediction accuracy of roughly 74–90% depending on the category.

2GIS

Pavel Denisov

Support Team Lead