diff --git a/src/pages/cloud/CloudUseCases.tsx b/src/pages/cloud/CloudUseCases.tsx
index c1b2802..2000e98 100644
--- a/src/pages/cloud/CloudUseCases.tsx
+++ b/src/pages/cloud/CloudUseCases.tsx
@@ -3,35 +3,33 @@ import { Eyebrow, SectionHeader, P, Small } from '../../components/Texts'
 
 const useCases = [
   {
-    title: 'AI / ML Training',
+    title: 'Enterprise & Private Kubernetes',
     description:
-      'Run GPU-accelerated workloads for deep learning, data science, and inference on demand.',
-    bullets: ['GPU acceleration', 'Scalable compute', 'Cost optimization'],
-  },
-  {
-    title: 'Enterprise Kubernetes',
-    description:
-      'Operate production-grade clusters with complete control and no vendor lock-in.',
-    bullets: ['High availability', 'Security', 'Compliance'],
-  },
-  {
-    title: 'Edge & IoT',
-    description:
-      'Leverage global nodes for low-latency workloads and connected device deployments.',
-    bullets: ['Low latency', 'Global distribution', 'Real-time processing'],
-  },
-  {
-    title: 'DigitalMe Blueprint',
-    description:
-      'Launch a full digital workspace on Mycelium Cloud with pre-integrated services.',
+      'Production-grade K3s clusters with full workload ownership and residency control. Deploy apps and internal services without vendor lock-in or exposed surfaces.',
     bullets: [
-      'Cryptpad • Encrypted collaboration',
-      'Elements • Matrix chat',
-      'Stallwart • Mail, calendar, contacts',
-      'Gitea • Git hosting',
-      'Nextcloud • File storage and sync',
-      'LiveKit / Jitsi • Real-time video',
-      'Single Sign-On backed by Gitea',
+      'High availability + failover',
+      'Encrypted, mesh-native networking',
+      'Compliance-ready governance & policy control',
+    ],
+  },
+  {
+    title: 'Edge & Distributed Deployments',
+    description:
+      'Place compute close to where data is created. Run clusters globally across home labs, offices, data centers, or edge locations.',
+    bullets: [
+      'Low-latency execution',
+      'Mesh-routed connectivity',
+      'Autonomous healing across region',
+    ],
+  },
+  {
+    title: 'AI / ML & Agent Workloads',
+    description:
+      'Run inference, training, and autonomous AI systems on sovereign compute. GPU orchestration, private vector storage, and model execution without handing data to hyperscalers.',
+    bullets: [
+      'Scales across nodes',
+      'Supports open-source & fine-tuned models',
+      'Built for agentic automation + RAG',
     ],
   },
 ]