2026年顶级数据趋势

Industry experts share theirpredictions for the year ahead Table of Contents Contributors ARMONPETROSSIAN BHASKARCHATURVEDIEnterprise Data Architect SATISHJAYANTHI LEE DERKSCEO GLEBMEZHANSKIY DAVID YAFFECo-Founder MATT FLORIANPractice Director -SAP Delivery ERIKDUFFIELD TEJASMANOHARCo-CEO & Co-Founder BARR MOSESCEO & Co-FounderMonte Carlo COLIN ZIMACEO & Co-FounderOmni FRANK BELLFounderData Thought Leader JOE REISAuthorData architect JAY GIMPLEChief Data &Analytics Officer MAC NOLANDChief Data Officer &Co-Founder CHRIS RÜGEHead of Data & BIRSG Group MIKE PALMERCEO JENNIFERBELISSENT ANDREW CRISPVP, Director ofEnterprise Data Services KENTGRAZIANO NICHOLASMANNFounder & CEO AI moves frompromise to proof 2026 will be the year of reckoning for AI initiatives. After several yearsof hype, experimentation, and inflated expectations, organizations willshift from proof-of-concept to proof-of-value. They will realize that AI With this shift comes both clarity and complexity. AI agents are evolvingfrom concept to reality, automating not just individual tasks but entireworkflows. Natural language is becoming the primary interface for datainteraction, opening up analytics capabilities to people who have never Yet alongside this progress comes a renewed focus on the fundamentals:data quality, governance, and modeling, which are proving more criticalthan ever. The companies pulling ahead aren’t those with perfect data— The insights shared in this year’s report come from data leaders who arenavigating this transformation firsthand, hard at work in the trenches ofreal AI implementation. Their predictions and observations reflect both The AI-powered datacore will be built on BHASKARCHATURVEDIEnterprise Open standards will unlock real-time,cross-cloud AI at scale Agentic AI emerges as the ETLand migration game-changer Let’s face it—AI is here to stay, and it’s starving for clean, unifieddata. Legacy silos are a death sentence for innovation, andthat’s why interoperable lakehouses are stealing the show.By 2026 and beyond, 60% of enterprises will lean on theseplatforms to merge structured and unstructured data, fueling AIwith real-time, cross-cloud access. Open standards like Apache Manual data migrations and clunky ETL pipelines are the stuff ofnightmares, slowing down modernization and racking up costs.Agentic AI is here to save the day, automating everything fromschema mapping to pipeline orchestration. By 2026, over 60% To capitalize on this shift, companies need to build lakehouseswith open table formats and federated querying, ensuring their The winning approach is to deploy AI-orchestrated pipelinesusing AI-powered DataOps to automate migrations with real The majority of businesses willquery data in plain English Imagine a world where anyone in your company can ask dataquestions and get instant, spot-on answers—no data scientistrequired. That’s where conversational AI is taking us. By2026, 70% of businesses will use NLP-driven platforms to letemployees query lakehouses and vector databases with plain Success here requires rolling out semantic layers with AI modelsthat turn natural language into real-time, context-rich insights, AI market hitshyperspeed, enterscorrection phase The M&A fever in the data space willcontinue as AI pressure mounts Competition will outpace Large companies will continue to realize the importance ofhaving a data strategy and scrambling to make moves intothe data space. We saw clear indicators of this in 2025 withSalesforce acquiring Informatica, ServiceNow acquiring data.world, and now Fivetran and dbt merging. For a company like AI is overshadowing everything right now, and agentic modelsand their impact are dominating the conversation. There’sdefinitely a bubble there, and while it’s hard to say if it’s going to What I’m really seeing is the rapid commoditization of AI. Thereare all these AI companies growing rapidly in the applicationlayers. As soon as one of them catches fire, they have twelvenew competitors within months, all doing the exact same things.It’s basically a market on hyperspeed—one player emerges, “The most exciting thing I see comingdown the road is agentic workflows indata to help accelerate use cases. I don’tthink that’s happened yet, but it’s closeand I think we’ll get there in 2026. In thefuture, nontechnical people could call The question is whether there’s a new stack emerging for AI.Everybody wants to believe that, but I think it will be made up ofexisting companies rather than new ones. I don’t see any new ARMON PETROSSIANCoalesce “Right now, everybody’sgot this glitzy, starry-eyedapproach to AI and how AI will eventually break thestranglehold of legacy ETL tools Right now, a lot of what AI is doing in data engineering is thingslike adding column descriptions or table-level descriptions.While this is helpful and shaves off a lot of time, it’s not a radicalshift or acceleration in development. But agentic workflowspromise