The race in enterprise AI isn't just about agent capabilities, but about owning the central dashboard where employees direct agents across all applications (Salesforce, Jira, etc.). Companies like OpenAI and Microsoft are vying to become this primary interface, controlling the customer relationship and relegating other apps to the background.
Related Insights
OpenAI's "Frontier" platform architecture reveals a strategy to insert layers of intelligence and action *on top of* existing enterprise systems of record (e.g., CRMs). This positions OpenAI to capture user value and relationships, reducing established SaaS players to commoditized data repositories or "dumb pipes."
AI's biggest enterprise impact isn't just automation but a complete replatforming of software. It enables a central "context engine" that understands all company data and processes, then generates dynamic user interfaces on demand. This architecture will eventually make many layers of the traditional enterprise software stack obsolete.
Frontier is designed to be a central hub for deploying and managing AI agents across enterprise systems. This positions OpenAI to become the primary user interface for work, potentially demoting established SaaS tools like CRMs to mere data repositories.
OpenAI's new platform, Frontier, is designed for building 'AI co-workers' that can access a company's various data sources and systems. This represents a strategic move beyond single-user chatbots toward an enterprise-grade orchestration layer for managing teams of interconnected AI agents.
The core conflict in AI is over who owns the user interface. Model makers like OpenAI aim for a universal 'big brain' agent that consumes data, while data platforms like Snowflake are building specialized agents on top of their proprietary data to avoid becoming commoditized data pipes.
Infrastructure built for app-to-app integration, like Salesforce's MuleSoft, is being repurposed to govern, orchestrate, and secure AI agents. This 'agent fabric' provides a foundational control plane for managing complex agentic workflows across the enterprise, extending the value of existing integration investments.
Legacy systems like CRMs will lose their central role. A new, dynamic 'agent layer' will sit above them, interpreting user intent and executing tasks across multiple tools. This layer, which collapses the distance between intent and action, will become the primary place where work gets done.
Instead of interacting with SaaS GUIs (like Greenhouse for hiring), users will interact with AI agents. These agents will directly manipulate the underlying system-of-record data, managing entire workflows from a simple conversation and making the traditional SaaS application redundant.
SaaS products like Salesforce won't be easily ripped out. The real danger is that new AI agents will operate across all SaaS tools, becoming the primary user interface and capturing the next wave of value. This relegates existing SaaS platforms to a lower, less valuable infrastructure layer.
The current market of specialized AI agents for narrow tasks, like specific sales versus support conversations, will not last. The industry is moving towards singular agents or orchestration layers that manage the entire customer lifecycle, threatening the viability of siloed, single-purpose startups.