Skip to content
Graph of Plan

Competitive: Big tech enters agent infrastructure

Microsoft, Google, and Amazon are all investing heavily in AI agent capabilities. They could launch competing infrastructure that commoditises our platform.

The Risk

We’re building agent execution infrastructure. The hyperscalers have:

  • Massive compute resources
  • Existing cloud platforms with enterprise relationships
  • AI models (GPT, Gemini, Bedrock) they can bundle
  • Distribution through existing developer ecosystems (VS Code, GCP, AWS)

If Microsoft launches “Azure Agent Spaces” or Google ships “Vertex AI Agents” with similar capabilities at lower prices, our differentiation erodes rapidly.

Specific Threats

  1. Bundled offerings: Agent infrastructure included free with existing cloud subscriptions
  2. Vertical integration: End-to-end stacks (model + compute + storage + IDE) that lock in developers
  3. Price competition: Loss-leader pricing that’s impossible to match
  4. Distribution advantage: Pre-installed in tools developers already use
  5. Enterprise relationships: CIOs buy from vendors they already trust

Mitigations

Product Differentiation

  • Opinionated over general: Our products solve specific problems (delivery control, purchasing automation) rather than offering generic agent compute. We’re applications, not infrastructure.
  • Domain depth: Murphy understands delivery methodology. P4gent understands procurement relationships. Big tech offers primitives, not solutions.
  • Cloudflare-native: Building on Cloudflare, not hyperscalers, gives us a differentiated cost structure and avoids direct competition.

Strategic Positioning

  • SME focus: Big tech optimises for enterprise. We can win SMEs who want solutions, not platforms.
  • Speed: We can ship faster. Big tech has coordination overhead.
  • Complementary: Our products can run on top of big tech infrastructure if needed—they’re not necessarily competitors.

Business Model

  • Application revenue: Pricing based on value delivered (delivery saved, costs reduced) rather than compute consumed
  • Vertical SaaS margins: Application-layer businesses command better margins than infrastructure

Residual Risk

We cannot prevent big tech from entering this market. Our defence is solving specific problems better than general-purpose tools can. If we’re just “agent compute,” we lose. If we’re “the tool that catches delivery slips before they happen,” we have a moat.

Probability: High (big tech is already moving into agent infrastructure) Impact: Medium (affects positioning more than viability) Mitigation effectiveness: Good (product differentiation is real)