Industry executives and experts share their predictions for 2026. Read them in this 18th annual VMblog.com series exclusive.
By Bianca Lewis, Executive Director, OpenSearch Software Foundation
AI has been a forcing function for innovation in nearly every industry over the past few years. Since technologies like search and observability are interconnected with AI, they are even more predisposed to change. This year, as generative AI becomes more reliable and enterprise-level AI adoption begins in earnest, search and observability are ripe for evolution, and we’ll see shifts in collecting, structuring, retrieving and monitoring data.
Here are three trends to look out for in the coming year:
Traditional search endures – thanks to, not in spite of – AI hype
On the surface, it may seem like AI has come to take the place of traditional search. Search engines are providing more and more generative AI summaries, and some may expect the shiny new technology to edge out long-standing alternatives. However, although the situation looks grim on the front end, the back end relies even more heavily on search. As search platforms increasingly incorporate AI, the need for reliable, data-backed retrieval has only grown.
To generate the summaries users now expect, AI systems must retrieve accurate, timely, and contextually relevant information. This requires robust retrieval pipelines that rely on traditional search techniques – like keyword indexing, filtering, and relevance scoring – to parse logs, match structured and unstructured data, and surface the most pertinent results. Instead of replacing traditional search, generative AI has made it a foundational necessity.
Bigger won’t always be better
Companies have spent millions of dollars procuring data, establishing partnerships with companies that have troves of content, and training AI models on swaths of material. But in 2026, bigger won’t always be better. Rather than just an expansive LLM, the key factor to building reliable AI will be data quality and data trust.
Tools like vector search and retrieval-augmented generation (RAG) are redefining AI reliability. Vector search identifies meaning-based connections in data, which is essential for nuanced queries, and RAG supplies real-time, context-rich material to ground AI responses. When combined, these enable systems to retrieve relevant, up-to-date information across multiple formats, like text, code, and logs, and reduce hallucinations. Leveraging these tools will ensure that information that powers generative AI and other AI applications is relevant and factually accurate. The largest LLM won’t always win. Instead, powering AI outputs with effective infrastructure and high-quality data will separate AI users from AI leaders.
A leaner future for AI and observability
Both the AI and observability markets are entering a consolidation phase. Recent moves – like Palo Alto Networks’s Chronosphere acquisition and LogicMonitor’s Catchpoint acquisition – signal a clear shift toward fewer, but more integrated solutions. Instead of stitching together separate tools for metrics, logs, traces and AI analytics, enterprises want unified platforms with built-in intelligence. Industry consolidation reflects a growing fatigue with the complexity and cost of disparate toolchains, and in 2026, organizations will increasingly favor fewer platforms that can do more. As search, observability, and AI all impact each other’s success, users will seek platforms that integrate those technologies to enhance all of their functionalities.
As these trends converge, the broader implications for infrastructure are becoming clear. 2026 will bring changes in how AI is built and deployed. Models aren’t the only factor anymore. Infrastructure will come into the spotlight as organizations realize the need for scalable, secure stacks that include indexing pipelines, hybrid search engines, and context-aware data integration across AI and observability workflows. Traditional technologies like semantic and keyword search will be adapted – not abandoned – and combined with modern capabilities like vector search, RAG, and unified observability. The result will be leaner, more reliable systems that enable real-time AI applications at enterprise scale. In this new era, success won’t be measured by how large your model is, but by how well your infrastructure can support it.
##
ABOUT THE AUTHOR
Bianca Lewis brings 20+ years of tech leadership and a track record of building growth around search and analytics to her role as executive director of the OpenSearch Software Foundation. As executive director, Bianca drives vendor-neutral, community-first development, strengthens collaboration across the ecosystem, and ensures the Foundation’s long-term sustainability. She also champions diversity in tech, serving on multiple advisory boards including LGBTWork. Bianca formerly served as chief revenue officer at Opster.
