chokepoints.ai
SUBSCRIBE
10 layers580 nodes2,376 dependencies95 chokepoints6,500+ companiesnode size = companies identified
← THE STACK

L8

Applications & Inference

This is where the AI supply chain meets paying customers. It covers the products and platforms that put AI in front of end users: consumer apps like ChatGPT, vertical software for healthcare, legal and finance, autonomous agents that complete multi-step tasks, and AI features embedded into incumbent platforms like Microsoft 365 and Salesforce. Everything above this layer was built in anticipation of demand here, which makes this layer the one investors across the whole chain are watching most closely.

L8INPUTSL8BUYERSAPPLICATIONS & INFERENCESUPPLY SCHEMATIC · L8

THE TAKE

This is where the revenue has to show up. A $690B+ build-out in 2026 is being funded ahead of demonstrated end-revenue, and the gap between those two numbers is the question every investor in the layers above has to answer.

Signals

  • Hyperscalers are tracking ~$725B of combined 2026 capex (+77% YoY) against far smaller demonstrated AI end-revenue (company filings).
  • Microsoft's AI business hit a ~$37B annualized run-rate, up ~123% YoY, with ~20M paid Copilot seats (company earnings).
  • Anthropic reached ~$30B ARR by April 2026, overtaking OpenAI's ~$25B (company / press).
  • Salesforce Agentforce ARR reached ~$800M (+169% YoY); Harvey hit ~$300M ARR at an ~$11B valuation (company filings / press).

The investment angle

The roughly 30:1 capex-to-revenue ratio is the central bull/bear pivot; firms that prove durable enterprise attach beyond developer tools command the safest inference demand.

Dominant playerVertical & agentic AI
ConcentrationDemand fragmented across software incumbents & startups
Key metricHyperscaler capex on track to exceed $690B+ in 2026
GeographyGlobal

Inside this layer, node by node

The atlas data behind this layer: 17 nodes, 0 of them chokepoints. Every node links back into the network map; market figures carry their source.

Agents and copilotsL8.3fragmentedscaling4 companies

Autonomous agents run multi-step tasks; copilots assist inside workflows. Both form the interface between models and enterprise operations. Incumbents dominate copilot distribution; point solutions remain fragmented.

$7.8B market · 202546.3% CAGRsource ↗
AI embedded in incumbent softwareL8.4oligopolyscaling7 companies

SaaS and enterprise platforms are adding generative and ML features to existing products. Their installed base and data gravity let them generate AI revenue faster than new entrants at scale. Established ISVs with distribution and customer lock-in capture most of this value.

$11B market · 202315.1% CAGRsource ↗

Companies we track

Microsoft
Copilot; ~$37B AI ARR; broadest distribution
MSFT · US
Anthropic
Claude; API-first enterprise layer
Private · US
OpenAI
ChatGPT; dominant consumer AI app
Private · US
Salesforce
Agentforce; CRM agentic workflows
CRM · US
ServiceNow
Now Assist; IT-ops agentic automation
NOW · US
Harvey
legal AI agents; vertical-SaaS bellwether
Private · US

Supply chain

Raw inputs

Frontier & open modelsUS
capability inputs

Key suppliers

Microsoft CopilotUS
incumbent distribution
Vertical AI (Harvey, Sierra)US
agentic & domain apps
ServiceNow / SalesforceUS
enterprise platforms

Buyers

Enterprises & consumersGlobal
end demand (the question)