25 companies most exposed to AI workload margin pressure

Horizontal SaaS (AI added to seat-based SaaS)

• Salesforce
• ServiceNow
• HubSpot
• Atlassian
• Monday.com

These companies integrate AI features into traditional SaaS pricing models, which historically produced very high margins but may face pressure as inference workloads scale.

AI Application Platforms

• Palantir
• C3.ai
• UiPath
• SoundHound AI
• BigBear.ai

These businesses run AI workloads directly inside their core products, meaning inference costs scale with usage.

DevOps / Observability Platforms

• Datadog
• Dynatrace
• Elastic
• GitLab
• JFrog

Observability platforms process large volumes of telemetry data to monitor systems and applications. Adding AI analysis increases compute demand on top of those workloads.

AI Workflow / Automation

• Pegasystems
• NICE
• Verint
• Automation Anywhere
• DataRobot

These platforms orchestrate enterprise workflows or AI agents that may require multiple model calls per task.

Edge / CDN / Inference Infrastructure

• Cloudflare
• Fastly
• Akamai
• Digital
• OceanVercel

Edge platforms run distributed compute close to users, which makes them natural hosts for inference workloads but also exposes them to compute cost scaling.