In a move challenging decades of software design, startup Eragon has secured $12 million in funding to pursue a radical vision: replacing traditional enterprise software interfaces with natural language prompts. Founded in August by former Oracle and Salesforce executive Josh Sirota, the company is building what it calls an "agentic AI operating system" for business.
The funding round, led by investors including Arielle Zuckerberg at Long Journey Ventures and Axiom Partners, values Eragon at $100 million post-money. Founder Josh Sirota declared "software is dead," arguing that graphical user interfaces are a bygone era. Eragon's platform aims to offer access to common business software suites—from Salesforce and Snowflake to Tableau and Jira—through a single large language model (LLM) interface.
Technical execution is led by a team including Berkeley and MIT PhDs, focusing on post-training open-source models like Qwen and Kimi on specific customer datasets. Crucially, the company emphasizes a security-first model where a client's data never leaves its own servers; AI models are trained and deployed within the customer's own secure cloud environment.
During a demonstration, Sirota showed the system onboarding a new customer via a simple prompt, which automatically assigned credentials, spun up a cloud instance, and initiated a tailored workflow. The platform can generate dashboards, analyze deal slippage, or suggest supply chain optimizations through conversational commands.
However, the vision faces challenges including AI hallucinations and auditability. The competitive landscape is intensifying, with Nvidia CEO Jensen Huang recently presenting a similar vision at the GTC conference, announcing the NemoClaw initiative for AI agents in secure enterprise systems.
Eragon is currently deployed in several large enterprises and dozens of smaller startups. Sirota frames this shift as analogous to the historical move from mainframes to personal computers, aiming to provide the foundational OS for specialized corporate agents.
