{"id":32767,"date":"2026-03-26T13:48:46","date_gmt":"2026-03-26T20:48:46","guid":{"rendered":"https:\/\/www.pingcap.com\/?p=32767"},"modified":"2026-03-27T13:49:10","modified_gmt":"2026-03-27T20:49:10","slug":"agentic-ai-systems-architecture-4-decisions-that-scale","status":"publish","type":"post","link":"https:\/\/www.pingcap.com\/ko\/blog\/agentic-ai-systems-architecture-4-decisions-that-scale\/","title":{"rendered":"Four Data Architecture Decisions That Make or Break Agentic Systems"},"content":{"rendered":"<p><em>Editor\u2019s note: This post originally appeared on The New Stack and is republished with permission. The original version is available <a href=\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/\">\uc5ec\uae30<\/a>.<\/em><\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Key_Takeaways\"><\/span><strong>Key Takeaways<\/strong><span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Agentic systems need architectures built for isolation, elasticity, and constant change.<\/li>\n\n\n\n<li>Separating compute from storage \u2014 and workloads from each other \u2014 is critical to scaling reliably.<\/li>\n\n\n\n<li>Cost visibility helps teams manage fast-growing agent workloads before spend gets out of control.<\/li>\n\n\n\n<li>Object storage and online change make it possible to recover, scale, and evolve without downtime.<\/li>\n<\/ul>\n<\/blockquote>\n\n\n\n<p>Data teams that thrived in the last wave of Software as a Service (SaaS)\u00a0platform scale weren\u2019t the ones that chased hype. They were the ones that made a few smart decisions: They adopted cloud-first operations, made cost and capacity visible, and chose architectures that could quickly adapt to changing conditions.\u00a0<\/p>\n\n\n\n<p>As it turns out, those are exactly the same practices the agentic era now demands.<\/p>\n\n\n\n<p>If you look at how data teams are managing their AI transition, a clear pattern emerges: The ones with firm control over performance and spend had the easiest time supporting agents. They were already practicing tenant isolation, making online changes during business hours, and using object-storage-backed recovery. Everything agents required was already in place. All they had to do was apply those same principles to a new kind of user.<\/p>\n\n\n\n<p>Let\u2019s start from that premise: that agents are your new users. We\u2019ll take a look at what makes them different from other users and how best to support them. Along the way, we\u2019ll discuss four architectural factors that affect the ability to operate at scale. And we\u2019ll close with a checklist you can use to evaluate the suitability of your current platform for agent-driven workloads.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Agents_Are_Your_New_%E2%80%98Users\"><\/span>Agents Are Your New \u2018Users\u2019<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p>Most data platforms were built for humans and services with relatively stable, predictable demands. Agentic systems are quite different. They spin up short-lived apps, run experiments, trigger migrations, branch off new datasets and tear it all down \u2014 often in parallel, and unpredictably.<\/p>\n\n\n\n<p>We\u2019ve seen this firsthand with companies like <a href=\"https:\/\/www.pingcap.com\/ko\/case-study\/manus-agentic-ai-database-tidb\/\">Manus<\/a>. It offers a general-purpose agentic AI platform whose \u201cwide research\u201d agent swarms spin up thousands of short-lived workloads every day. It\u2019s&nbsp; no longer managing a single monolithic database, instead&nbsp; orchestrating millions of tiny, temporary branch-like environments behind the scenes.<\/p>\n\n\n\n<p>At scale, what agents need isn\u2019t a monolithic, ever-growing database. It\u2019s effectively millions of tiny, isolated databases or branches popping in and out of existence. Once you accept that premise, four requirements for agentic architectures naturally follow:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Isolation by default:<\/strong> Per-tenant or per-agent boundaries keep experiments from becoming everybody\u2019s problem.<\/li>\n\n\n\n<li><strong>Online change during business hours:<\/strong> Schemas and indexes must be adjustable within p95\/p99 latencies inside service-level objectives (SLOs).<\/li>\n\n\n\n<li><strong>Placement and quotas:<\/strong> Hot data needs to be kept near low-latency compute; cold data can be kept in cheap storage; and noisy tenants need to be isolated.<\/li>\n\n\n\n<li><strong>Life cycle automation:<\/strong> You need the ability to create and retire environments in seconds with clean metadata hygiene and cost attribution.<\/li>\n<\/ul>\n\n\n\n<p>Here are four architectural choices that support your agentic users\u2019 needs.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"1_The_Two_Separations_That_Matter\"><\/span>1. The Two Separations That Matter<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p>Agents can quickly devour shared resources. To keep that from happening, separate compute from storage so you can add query capacity without shifting data. Then separate compute from compute to give online transaction processing (OLTP), analytics and maintenance their own lanes and SLOs.&nbsp;<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Separate Compute from Storage<\/h3>\n\n\n\n<p>Attaching stateless SQL\/compute engines to durable, shared object storage lets you:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Scale elastically:<\/strong> Adding and removing query capacity without the need for high-wire data copies or weekend migrations.<\/li>\n\n\n\n<li><strong>Recover predictably:<\/strong> New nodes can pull state from storage and warm caches and start serving without saturating peers.<\/li>\n\n\n\n<li><strong>Clone quickly:<\/strong> Copy-on-write branches can be built quickly from metadata rather than complete physical copies.<\/li>\n<\/ul>\n\n\n\n<p>What to verify:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Can you add compute nodes in minutes without rebalancing data?<\/li>\n\n\n\n<li>Do new nodes draw from object storage rather than peers?<\/li>\n\n\n\n<li>Are cloning and branching incremental and space-efficient?<\/li>\n<\/ul>\n\n\n\n<h3 class=\"wp-block-heading\">Separate Compute from Compute<\/h3>\n\n\n\n<p>When thousands of agents are branching data, building indexes and sending queries at the same time, SQL frontends, analytical readers, background maintenance (compaction, backfills), backup\/restore and control planes need to be scaled \u2014 and governed \u2014 independently, to keep them running smoothly.<\/p>\n\n\n\n<p>What to verify:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Can you rate-limit backfills independently of OLTP traffic?<\/li>\n\n\n\n<li>Do analytical scans have their own resources and guardrails?<\/li>\n\n\n\n<li>Can you execute version upgrades for one plane without taking a window on another?<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"2_Make_Cost_Visible_and_Actionable\"><\/span>2. Make Cost Visible (and Actionable)<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p>Traditional data platforms often idle at 20 to 25% CPU while maintaining extra headroom \u201cjust in case.\u201d That\u2019s survivable with human users; it\u2019s untenable in an environment where agents are spinning up thousands of short-lived workloads. The fix is to make the cost per query visible \u2014 for example, through request-unit (RU) accounting \u2014 in the same pane engineers already watch.<\/p>\n\n\n\n<p>That way, engineers know which queries to optimize and what savings to expect. Product and finance can set budgets and caps that map to real work, and platform teams can recommend improvements based on actual spend, not gut feel.<\/p>\n\n\n\n<p>What to verify:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Can you attribute costs to tenants, apps and query digests?<\/li>\n\n\n\n<li>Can you enforce&nbsp; budgets and caps automatically?<\/li>\n\n\n\n<li>Do you have a \u201cTop Five Digests\u201d loop tied to latency and cost regression tests?<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"3_Treat_Object_Storage_as_the_Backbone\"><\/span>3. Treat Object Storage as the Backbone<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p>For agentic architectures, using object storage (S3\/Google Cloud Storage\/Azure Blob) for the data backbone is not optional. It enables context-aware scaling by pulling data from a shared object store and caching hot data locally for ultra-low latency, ensuring the database is always the right size for the moment. During scale-out or recovery, new compute should pull state from durable storage rather than copying from peers. Backups and long-term snapshots should live there too.<\/p>\n\n\n\n<p>Benefits:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Predictable scale and recovery: Less cross-node thrash during growth or failover.<\/li>\n\n\n\n<li>Tiered economics<strong>:<\/strong> Hot\/warm\/cold paths you can reason about and budget for.<\/li>\n\n\n\n<li>Fast database branching: Database clones become pointer operations plus object-store semantics.<\/li>\n<\/ul>\n\n\n\n<p>What to verify:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Are backups, snapshots and branch metadata stored in object storage by default?<\/li>\n\n\n\n<li>How long does it take for a new node to start serving traffic after a failure?<\/li>\n\n\n\n<li>Can you garbage-collect abandoned branches and objects automatically?<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"4_Treat_Online_Change_as_a_First-Class_Capability\"><\/span>4. Treat Online Change as a First-Class Capability<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p>When agents are your users, change is constant. Schema evolution, indexing, data movement and upgrades must happen online, with clear visibility into what is happening.<\/p>\n\n\n\n<p>Here\u2019s what that looks like in practice:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Three-phase schema changes (prepare \u2192 reorganize \u2192 commit) with multiversion concurrency control so reads\/writes continue while backfills run.<\/li>\n\n\n\n<li>Rate-limited maintenance that respects p95\/p99 budgets.<\/li>\n\n\n\n<li>Rolling upgrades with automatic leader election and no maintenance windows.<\/li>\n<\/ul>\n\n\n\n<p>What to verify:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Can you add an index to a hot table at peak and hold p95\/p99 inside the SLO?<\/li>\n\n\n\n<li>Are metadata locks short and predictable?<\/li>\n\n\n\n<li>Do you have preflight checks, abort thresholds and a rollback plan baked into the pipeline?<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Anti-Patterns_to_Avoid\"><\/span>Anti-Patterns to Avoid<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p>So that\u2019s what you should try to do. Here are some things to avoid.<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Sharding complexity:<\/strong> App-level sharding looks simple until you own routing, rebalancing, failover and cross-shard joins forever.<\/li>\n\n\n\n<li><strong>One big pool:<\/strong> Treating all compute as fungible leads to noisy-neighbor incidents and tail-latency spikes.<\/li>\n\n\n\n<li><strong>Invisible spend:<\/strong> Billing at the instance level hides per-query waste; remember you can\u2019t manage what you can\u2019t see.<\/li>\n\n\n\n<li><strong>Peer copy dependency:<\/strong> Recovery and scale-out processes that depend on busy neighbors are vulnerable to collapsing under pressure.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"A_Minimal_Evaluation_Checklist\"><\/span>A Minimal Evaluation Checklist<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p>Use the following checklist to compare platforms for agentic workloads:<\/p>\n\n\n\n<ol class=\"wp-block-list\">\n<li><strong>Database provisioning:<\/strong> How many isolated databases, schemas and branches can you create per minute? How are they tracked and retired?<\/li>\n\n\n\n<li><strong>Two separations:<\/strong> Check compute\/storage independence and compute\/compute independence under live load.<\/li>\n\n\n\n<li><strong>Cost model:<\/strong> How well can engineers monitor per-query cost by tenant\/app? What caps exist and how are they enforced?<\/li>\n\n\n\n<li><strong>Object storage:<\/strong> Demonstrate node join and recovery that draw from object storage. Measure time to service.<\/li>\n\n\n\n<li><strong>Online change:<\/strong> Test the ability to add an index during peak; check p95\/p99, error rates and abort thresholds.<\/li>\n\n\n\n<li><strong>Failure drill:<\/strong> Kill a leader or availability zone (AZ); watch election, client retries and tail latency.<\/li>\n\n\n\n<li><strong>Metadata hygiene:<\/strong> Prove that abandoned branches and objects get garbage collected without manual tickets.<\/li>\n<\/ol>\n\n\n\n<p>Agentic systems don\u2019t require a brand new approach to infrastructure. The right architecture for agents is the right architecture for any large-scale modern use case. But agents are<em> <\/em>a forcing function.&nbsp;<\/p>\n\n\n\n<p>Data teams don\u2019t have the luxury of sticking with monolithic platforms that are slow to scale and hard to manage. Agents will bring those old architectures to their knees. But as the most successful data teams have found, if you design for flexibility, visibility and performance using the methods described above, you\u2019ll ship faster with fewer weekend fire drills, even when your \u201cusers\u201d number in the millions \u2014 and most of them aren\u2019t human.<\/p>\n\n\n\n<p><em>Ready to evaluate your architecture for agentic workloads? <a href=\"http:\/\/Ready to evaluate your architecture for agentic workloads? See how PingCAP helps teams build and scale AI agents with isolated database environments.\">See how TiDB helps teams build and scale AI agents with isolated database environments<\/a>.<\/em><\/p>","protected":false},"excerpt":{"rendered":"<p>Editor\u2019s note: This post originally appeared on The New Stack and is republished with permission. The original version is available here. Data teams that thrived in the last wave of Software as a Service (SaaS)\u00a0platform scale weren\u2019t the ones that chased hype. They were the ones that made a few smart decisions: They adopted cloud-first [&hellip;]<\/p>\n","protected":false},"author":50,"featured_media":32772,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"ub_ctt_via":"","footnotes":""},"categories":[145],"tags":[441,28,482,9,111],"class_list":["post-32767","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-thought-leadership","tag-agentic-ai","tag-architecture","tag-object-storage","tag-scalability","tag-tidb"],"acf":[],"featured_image_src":"https:\/\/static.pingcap.com\/files\/2026\/03\/27133857\/tidb_feature_1800x600-1-4.png","author_info":{"display_name":"Max Liu","author_link":"https:\/\/www.pingcap.com\/ko\/blog\/author\/max-liu\/"},"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v26.9 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Agentic AI Systems Architecture: 4 Decisions That Scale<\/title>\n<meta name=\"description\" content=\"Learn four agentic AI systems architecture decisions for scaling AI agents: isolation, cost visibility, object storage, and online change.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/\" \/>\n<meta property=\"og:locale\" content=\"ko_KR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Agentic AI Systems Architecture: 4 Decisions That Scale\" \/>\n<meta property=\"og:description\" content=\"Learn four agentic AI systems architecture decisions for scaling AI agents: isolation, cost visibility, object storage, and online change.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/\" \/>\n<meta property=\"og:site_name\" content=\"TiDB\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/facebook.com\/pingcap2015\" \/>\n<meta property=\"article:published_time\" content=\"2026-03-26T20:48:46+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-03-27T20:49:10+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/static.pingcap.com\/files\/2026\/03\/27132731\/tidb_1200x627-1-2.png\" \/>\n\t<meta property=\"og:image:width\" content=\"2400\" \/>\n\t<meta property=\"og:image:height\" content=\"1254\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"Max Liu\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:image\" content=\"https:\/\/static.pingcap.com\/files\/2026\/03\/27132746\/tidb_twitter_1600x900-4.png\" \/>\n<meta name=\"twitter:creator\" content=\"@PingCAP\" \/>\n<meta name=\"twitter:site\" content=\"@PingCAP\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Max Liu\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"7\ubd84\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/www.pingcap.com\/blog\/agentic-ai-systems-architecture-4-decisions-that-scale\/\"},\"author\":{\"name\":\"Max Liu\",\"@id\":\"https:\/\/www.pingcap.com\/#\/schema\/person\/7aeccb8936b7f908f923bdd0a6b44be8\"},\"headline\":\"Four Data Architecture Decisions That Make or Break Agentic Systems\",\"datePublished\":\"2026-03-26T20:48:46+00:00\",\"dateModified\":\"2026-03-27T20:49:10+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/www.pingcap.com\/blog\/agentic-ai-systems-architecture-4-decisions-that-scale\/\"},\"wordCount\":1508,\"publisher\":{\"@id\":\"https:\/\/www.pingcap.com\/#organization\"},\"image\":{\"@id\":\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/static.pingcap.com\/files\/2026\/03\/27133857\/tidb_feature_1800x600-1-4.png\",\"keywords\":[\"Agentic AI\",\"Architecture\",\"Object Storage\",\"Scalability\",\"TiDB\"],\"articleSection\":[\"Thought Leadership\"],\"inLanguage\":\"ko-KR\"},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.pingcap.com\/blog\/agentic-ai-systems-architecture-4-decisions-that-scale\/\",\"url\":\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/\",\"name\":\"Agentic AI Systems Architecture: 4 Decisions That Scale\",\"isPartOf\":{\"@id\":\"https:\/\/www.pingcap.com\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/static.pingcap.com\/files\/2026\/03\/27133857\/tidb_feature_1800x600-1-4.png\",\"datePublished\":\"2026-03-26T20:48:46+00:00\",\"dateModified\":\"2026-03-27T20:49:10+00:00\",\"description\":\"Learn four agentic AI systems architecture decisions for scaling AI agents: isolation, cost visibility, object storage, and online change.\",\"breadcrumb\":{\"@id\":\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#breadcrumb\"},\"inLanguage\":\"ko-KR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"ko-KR\",\"@id\":\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#primaryimage\",\"url\":\"https:\/\/static.pingcap.com\/files\/2026\/03\/27133857\/tidb_feature_1800x600-1-4.png\",\"contentUrl\":\"https:\/\/static.pingcap.com\/files\/2026\/03\/27133857\/tidb_feature_1800x600-1-4.png\",\"width\":1800,\"height\":600},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/www.pingcap.com\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Four Data Architecture Decisions That Make or Break Agentic Systems\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.pingcap.com\/#website\",\"url\":\"https:\/\/www.pingcap.com\/\",\"name\":\"TiDB\",\"description\":\"TiDB | SQL at Scale\",\"publisher\":{\"@id\":\"https:\/\/www.pingcap.com\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.pingcap.com\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"ko-KR\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/www.pingcap.com\/#organization\",\"name\":\"PingCAP\",\"url\":\"https:\/\/www.pingcap.com\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ko-KR\",\"@id\":\"https:\/\/www.pingcap.com\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/static.pingcap.com\/files\/2021\/11\/pingcap-logo.png\",\"contentUrl\":\"https:\/\/static.pingcap.com\/files\/2021\/11\/pingcap-logo.png\",\"width\":811,\"height\":232,\"caption\":\"PingCAP\"},\"image\":{\"@id\":\"https:\/\/www.pingcap.com\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/facebook.com\/pingcap2015\",\"https:\/\/x.com\/PingCAP\",\"https:\/\/linkedin.com\/company\/pingcap\",\"https:\/\/youtube.com\/channel\/UCuq4puT32DzHKT5rU1IZpIA\"]},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.pingcap.com\/#\/schema\/person\/7aeccb8936b7f908f923bdd0a6b44be8\",\"name\":\"Max Liu\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ko-KR\",\"@id\":\"https:\/\/www.pingcap.com\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/static.pingcap.com\/files\/2022\/10\/17234942\/avatar.jpg\",\"contentUrl\":\"https:\/\/static.pingcap.com\/files\/2022\/10\/17234942\/avatar.jpg\",\"caption\":\"Max Liu\"},\"description\":\"Co-Founder &amp; CEO\",\"sameAs\":[\"https:\/\/www.linkedin.com\/in\/max-liu-0b2390147\/\"],\"url\":\"https:\/\/www.pingcap.com\/ko\/blog\/author\/max-liu\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Agentic AI Systems Architecture: 4 Decisions That Scale","description":"Learn four agentic AI systems architecture decisions for scaling AI agents: isolation, cost visibility, object storage, and online change.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/","og_locale":"ko_KR","og_type":"article","og_title":"Agentic AI Systems Architecture: 4 Decisions That Scale","og_description":"Learn four agentic AI systems architecture decisions for scaling AI agents: isolation, cost visibility, object storage, and online change.","og_url":"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/","og_site_name":"TiDB","article_publisher":"https:\/\/facebook.com\/pingcap2015","article_published_time":"2026-03-26T20:48:46+00:00","article_modified_time":"2026-03-27T20:49:10+00:00","og_image":[{"width":2400,"height":1254,"url":"https:\/\/static.pingcap.com\/files\/2026\/03\/27132731\/tidb_1200x627-1-2.png","type":"image\/png"}],"author":"Max Liu","twitter_card":"summary_large_image","twitter_image":"https:\/\/static.pingcap.com\/files\/2026\/03\/27132746\/tidb_twitter_1600x900-4.png","twitter_creator":"@PingCAP","twitter_site":"@PingCAP","twitter_misc":{"Written by":"Max Liu","Est. reading time":"7\ubd84"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#article","isPartOf":{"@id":"https:\/\/www.pingcap.com\/blog\/agentic-ai-systems-architecture-4-decisions-that-scale\/"},"author":{"name":"Max Liu","@id":"https:\/\/www.pingcap.com\/#\/schema\/person\/7aeccb8936b7f908f923bdd0a6b44be8"},"headline":"Four Data Architecture Decisions That Make or Break Agentic Systems","datePublished":"2026-03-26T20:48:46+00:00","dateModified":"2026-03-27T20:49:10+00:00","mainEntityOfPage":{"@id":"https:\/\/www.pingcap.com\/blog\/agentic-ai-systems-architecture-4-decisions-that-scale\/"},"wordCount":1508,"publisher":{"@id":"https:\/\/www.pingcap.com\/#organization"},"image":{"@id":"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#primaryimage"},"thumbnailUrl":"https:\/\/static.pingcap.com\/files\/2026\/03\/27133857\/tidb_feature_1800x600-1-4.png","keywords":["Agentic AI","Architecture","Object Storage","Scalability","TiDB"],"articleSection":["Thought Leadership"],"inLanguage":"ko-KR"},{"@type":"WebPage","@id":"https:\/\/www.pingcap.com\/blog\/agentic-ai-systems-architecture-4-decisions-that-scale\/","url":"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/","name":"Agentic AI Systems Architecture: 4 Decisions That Scale","isPartOf":{"@id":"https:\/\/www.pingcap.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#primaryimage"},"image":{"@id":"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#primaryimage"},"thumbnailUrl":"https:\/\/static.pingcap.com\/files\/2026\/03\/27133857\/tidb_feature_1800x600-1-4.png","datePublished":"2026-03-26T20:48:46+00:00","dateModified":"2026-03-27T20:49:10+00:00","description":"Learn four agentic AI systems architecture decisions for scaling AI agents: isolation, cost visibility, object storage, and online change.","breadcrumb":{"@id":"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#breadcrumb"},"inLanguage":"ko-KR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/"]}]},{"@type":"ImageObject","inLanguage":"ko-KR","@id":"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#primaryimage","url":"https:\/\/static.pingcap.com\/files\/2026\/03\/27133857\/tidb_feature_1800x600-1-4.png","contentUrl":"https:\/\/static.pingcap.com\/files\/2026\/03\/27133857\/tidb_feature_1800x600-1-4.png","width":1800,"height":600},{"@type":"BreadcrumbList","@id":"https:\/\/thenewstack.io\/4-data-architecture-decisions-that-make-or-break-agentic-systems\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.pingcap.com\/"},{"@type":"ListItem","position":2,"name":"Four Data Architecture Decisions That Make or Break Agentic Systems"}]},{"@type":"WebSite","@id":"https:\/\/www.pingcap.com\/#website","url":"https:\/\/www.pingcap.com\/","name":"\ud2f0DB","description":"TiDB | SQL at Scale","publisher":{"@id":"https:\/\/www.pingcap.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.pingcap.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"ko-KR"},{"@type":"Organization","@id":"https:\/\/www.pingcap.com\/#organization","name":"PingCAP","url":"https:\/\/www.pingcap.com\/","logo":{"@type":"ImageObject","inLanguage":"ko-KR","@id":"https:\/\/www.pingcap.com\/#\/schema\/logo\/image\/","url":"https:\/\/static.pingcap.com\/files\/2021\/11\/pingcap-logo.png","contentUrl":"https:\/\/static.pingcap.com\/files\/2021\/11\/pingcap-logo.png","width":811,"height":232,"caption":"PingCAP"},"image":{"@id":"https:\/\/www.pingcap.com\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/facebook.com\/pingcap2015","https:\/\/x.com\/PingCAP","https:\/\/linkedin.com\/company\/pingcap","https:\/\/youtube.com\/channel\/UCuq4puT32DzHKT5rU1IZpIA"]},{"@type":"Person","@id":"https:\/\/www.pingcap.com\/#\/schema\/person\/7aeccb8936b7f908f923bdd0a6b44be8","name":"Max Liu","image":{"@type":"ImageObject","inLanguage":"ko-KR","@id":"https:\/\/www.pingcap.com\/#\/schema\/person\/image\/","url":"https:\/\/static.pingcap.com\/files\/2022\/10\/17234942\/avatar.jpg","contentUrl":"https:\/\/static.pingcap.com\/files\/2022\/10\/17234942\/avatar.jpg","caption":"Max Liu"},"description":"Co-Founder &amp; CEO","sameAs":["https:\/\/www.linkedin.com\/in\/max-liu-0b2390147\/"],"url":"https:\/\/www.pingcap.com\/ko\/blog\/author\/max-liu\/"}]}},"grav_blocks":false,"card_markup":"<a class=\"card-resource bg-white\" href=\"https:\/\/www.pingcap.com\/ko\/blog\/agentic-ai-systems-architecture-4-decisions-that-scale\/\"><div class=\"card-resource__image-container\"><img class=\"card-resource__image\" alt=\"tidb_feature_1800x600 (1)\" src=\"https:\/\/static.pingcap.com\/files\/2026\/03\/27133857\/tidb_feature_1800x600-1-4.png\" loading=\"lazy\" width=1800 height=600 \/><\/div><div class=\"card-resource__content-container\"><div class=\"card-resource__content-head\"><div class=\"card-resource__category\">Thought Leadership<\/div><\/div><h5 class=\"card-resource__title\">Four Data Architecture Decisions That Make or Break Agentic Systems<\/h5><\/div><\/a>","_links":{"self":[{"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/posts\/32767","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/users\/50"}],"replies":[{"embeddable":true,"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/comments?post=32767"}],"version-history":[{"count":5,"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/posts\/32767\/revisions"}],"predecessor-version":[{"id":32775,"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/posts\/32767\/revisions\/32775"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/media\/32772"}],"wp:attachment":[{"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/media?parent=32767"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/categories?post=32767"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.pingcap.com\/ko\/wp-json\/wp\/v2\/tags?post=32767"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}