Breaking down the tiers: How generative AI and knowledge-centered service are streamlining customer support

Automation, speed, and scale are at the core of generative AI’s ability to flatten the traditional support structure by eliminating the lower support tiers. Instead of routing cases through multiple tiers, generative AI can surface accurate answers instantly through natural language prompts — both for customers and support engineers. 

By implementing KCS and consolidating knowledge documentation and articles within a centralized knowledge base, we can empower customers to self-serve more effectively, reducing the volume of basic tier one cases. At the same time, support engineers gain faster access to contextual answers, lowering the need for internal escalations. This shift minimizes the operational reliance on lower-tier support and frees up high-value technical experts to focus on more complex, strategic work.

Customer support is always on, capable of handling thousands of interactions simultaneously without sacrificing quality. Ultimately, generative AI and KCS give customer support that’s more than just fast — it’s relevant, it’s personalized, and it consistently provides better outcomes.

I would be remiss not to highlight the hero that makes this improved customer support experience possible: unstructured data. Often, the data required to support KCS is not organized. Rather, it’s scattered across disparate sources in various formats such as tech docs, white papers, knowledge bases, case history resolutions, and more. This qualitative, unstructured data often contains rich insights that aren’t available in structured data. Managing this data effectively creates more opportunities for support engineers to enhance customer experience.

By integrating the data with semantic search, vector search, machine learning, and natural language processing (NLP), support teams can find results based on meaning.  Combining this with a large language model (LLM) enables generative AI to deliver tangible value by triggering real-time answer generation, intelligent case routing, and even proactive issue resolution.

Customer support has traditionally been reactive, responding only when something breaks or a case is submitted. But what if support could be more like preventive medicine: identifying risks early and intervening before problems escalate? Not only does it lead to higher customer satisfaction, but it also lowers organizational costs. 

By integrating with telemetry and behavioral analytics, AI can detect early warning signs and identify unusual patterns to predict issues before they arise. 

The next step? Teams engage customers proactively, resolving concerns before they become problems. They reduce the need for traditional support requests altogether. Just like that, generative AI enables support teams to transition from firefighting to autonomous stability management.

What does all of this mean for the role of the support? Like many professions, the support role is evolving in the face of GenAI. Representatives and engineers are no longer just troubleshooters — they are becoming strategic advisors, complex problem-solvers, and customer advocates. GenAI doesn’t replace their expertise; it amplifies it.

Modernizing support with GenAI raises natural questions around job security and changing responsibilities. But the reality is that automation frees engineers from repetitive, manual work — like digging through documentation or summarizing cases — and gives them back time to focus on deep, complex cases that require human ingenuity. This shift not only reduces burnout but creates space for more fulfilling, value-adding work.

One of the most meaningful changes is that the percentage of knowledge work is increasing. Engineers are no longer just consuming knowledge — they’re training and shaping GenAI systems by feeding them rich insights from real-world cases, especially edge cases or product-specific nuances. Their work directly impacts the quality and accuracy of AI responses downstream.

At the same time, GenAI proficiency is quickly becoming a core skill. Teams and individuals who adopt it effectively are already seeing tangible improvements in key support KPIs — including faster resolution times, increased deflection, and higher customer satisfaction scores. This means that the future belongs not just to support engineers, but to support engineers who use GenAI well.

We're also seeing a shift in team structure. Some headcount is being redirected to AI-focused roles — like data scientists, AI engineers, and GenAI specialists — who help build and evolve targeted systems that drive automation, self-service, and contextual support. These new roles complement engineering teams, enabling them to deliver even more value at scale.

Finally, with GenAI handling much of the high-volume, repetitive inbound work, engineers can engage more directly with customers — offering best practices, health checks, proactive recommendations, and consultative support. This personalized engagement not only improves outcomes but elevates support’s position within the business as a value-driving function.

To help teams thrive in this new environment, support leaders should communicate clearly about AI’s role, phase in adoption gradually, and invest in training to ensure every team member can thrive in an AI-augmented workflow.

At Elastic, we’ve created a GenAI-powered tool for our customers and support engineers. The Elastic Support Assistant — a generative AI application built on the Search AI Platform — empowers support engineers to efficiently respond to customer support issues and enables customers to self-serve via a chatbot experience. It helps customers answer their queries quickly and frees up engineers’ time.

Learn how we created the Support Assistant and what results we’ve seen so far. Read (and watch) the case study.