Cohesity

Solution

Cohesity built its NextGen Support Portal as an in-house, production-grade agentic AI support system for enterprise customers. The portal serves as the conversational front end where users describe issues in plain language instead of filling rigid forms. Inside that experience, a guided AI troubleshooting engine handles interactive diagnosis by asking clarifying questions, refining understanding, and leading customers through structured resolution steps. A retrieval augmented generation (RAG) engine indexes Cohesity’s knowledge base, product documentation, and historical case data so the system can generate contextually accurate, real-time responses grounded in support content. An agentic workflow layer then orchestrates multi-step actions such as interpreting problems, searching for relevant solutions, presenting options, confirming intent, and automatically creating support cases with populated priority, subject, description, tags, triage context, and suggested next steps for engineers.

Data flow

Customers enter a problem in the NextGen Support Portal in plain language. The guided AI troubleshooting engine engages the user, while the RAG engine retrieves context from the knowledge base, product documentation, and historical case data to generate answers. The agentic layer then guides resolution or, if escalation is needed, creates a support case pre-populated with conversation context, triage details, and suggested next steps for support engineers.