Maya Enterprise · Case study for technology industry · 2026
Novara Technologies: improving ROI in agentic workflows
Human Context Infrastructure for Agentic Workflows in Enterprise Technology. United Kingdom, 2026.
Novara Technologies is a 1,300 person UK enterprise SaaS company. As the independent testing agent, Takara.ai ran a standardized test across 50 employees representing a customer service team, 360 employee by task Work Contracts and 42 test cases across seven customer service task categories. With Maya, correction rounds fell from 3 to 2, and the agent followed more of the rules governing ownership, review, escalation and deadlines.
less rework. The back and forth between employee and agent fell from 3 correction rounds to 2
of the agreed rules followed by the agent, up from 32% without Maya
biggest gain in Service Requests: agreed rules followed rose from 25% to 83%
13 pages. Prepared by WORK·SELF, which builds Maya Enterprise, Human Context Infrastructure for enterprise AI.
