Silicon Valley's famous incubator Y Combinator (YC) publishes a "Request for Startups", outlining their anticipated entrepreneurial directions. In Fall 2025, YC listed five major tracks and one grand vision:
- AI Education: Enable AI to autonomously train technical workers and manual laborers
- AI Video: Reshape entertainment, e-commerce, and gaming industries with AI video technology
- AI Infrastructure: AI infrastructure supporting multi-agent systems
- AI Enterprise Services: Build native AI enterprise software systems
- AI Consulting: Enable large models to perform government certification, review, and consulting
And one grand vision: birth the first AI company with 10 people and a $100 billion valuation.
Among the five directions, I'm currently most involved in AI Video. But this article wants to systematically discuss the AI Enterprise Software Systems track that I've been thinking about for half a year. In my view, this is very likely the direction that will give birth to the first $100 billion AI company.
YC's analogy for AI-native enterprise software is: "Cursor for sales, HR, finance, and other departments." This metaphor is very vivid. The core idea is that AI enterprise operating systems should integrate management, business, knowledge, and all processes into an automated system, enabling AI and humans to collaborate efficiently. Currently, there are already many pioneering products that have simplified some business processes through AI integration, achieving great success:
- Vercel + Next.js templates: Partial automation of frontend development and operations, greatly reducing business complexity
- AI marketing tools represented by K:AI: Automating planning, execution, evaluation, and customer service processes
Although these products have shown the embryonic form of AI-integrated enterprise processes, they are mostly partial implementations. A true AI enterprise operating system needs to run through the entire enterprise lifecycle, connecting all links. Taking SaaS (subscription software) companies as an example, organizational structures typically include:
- Functional departments: Finance, HR, Administration, Legal, IT
- Business departments: Product, Design, Technology (frontend/backend), Operations, Testing, Sales, Customer Service, Marketing, Operations
Below, let's envision what a complete AI enterprise software system might look like. Suppose this AI system interface resembles Feishu or Slack, it not only supports traditional instant messaging and document collaboration but also embeds "AI employees." These AIs continuously read documents and messages, automatically responding with built-in business processes. Actual business scenarios might include:
- Product Manager: Use AI to analyze requirements, market, and product direction, automatically generate roadmaps, wireframes, documents, and iterate optimizations
- Project Manager: AI assists with project planning, task allocation, progress tracking, automatically completing subtasks
- Designer: Tools like Figma, Lovart combined with AI to generate and optimize design drafts
- Frontend/Backend Engineers: Use Cursor, Claude Code, V0, etc. for code generation and optimization, automatically maintain API documentation, achieve frontend-backend integration
- Testing: AI automatically generates and executes test cases, supports A/B testing, etc.
- Operations: AI Ops tools achieve full automation of deployment, monitoring, alerting, and log analysis
AI employees can leverage complete enterprise records to actively complete and maintain tasks and documents. When encountering difficult problems, they can proactively seek assistance from human colleagues, maintaining complete information context throughout the process.
In sales and operations, AI's role is equally prominent:
- Sales: AI CRM (Customer Relationship Management), automatically analyze customer conversations/emails and sentiment, recommend opportunities and action plans
- Customer Service: Intelligently categorize and respond to user questions, coordinate ticket creation and distribution, collaborate with AI developers to quickly provide solutions
- Operations: Automate planning and execution of operational activities, integrate with mainstream social media, analyze data, optimize strategies, covering multi-touchpoint operations like email/affiliate/KOL
- Advertising: Automatically generate copy, images, and videos based on needs, and efficiently place advertisements
A complete AI enterprise system can connect the business closed-loop from product to sales, achieving extreme efficiency from idea to first revenue within 24 hours.
Further, functional departments can also benefit:
- Finance: Full automation of financial reports, budgets, costs, and tax filing, real-time recording of information expenses, decision-makers can access operational data anytime
- HR: AI generates recruitment, onboarding, transfer, and departure processes, automatic performance evaluation, assists in resume screening and onboarding, generates training documents to help newcomers quickly adapt
- IT Department: AI maintains CRM, ERM, CMDB, SOP, automatically assigns permissions
- Administration: Automatically complete meeting room booking, travel, procurement, reimbursement, and other trivial matters
- Strategic Level: Based on complete historical data, AI assists strategic planning, market and competitive analysis, aids management in efficient decision-making
Through this system, advanced enterprise management models can be directly embedded into daily enterprise operations, and all employee work processes will be completely recorded. This not only helps continuously optimize AI-driven processes but also achieves precise quantification and incentives for human employee contributions. Enterprises only need to hire humans for positions where AI cannot perform, AI employees and human employees can collaborate complementarily, leveraging sufficient contextual information to maximize both parties' potential.
In summary, I believe the core value brought by AI-native enterprise software lies in:
- Reducing the cost of launching new businesses, improving enterprise leverage
- Precisely evaluating human employee contributions, providing more scientific incentive mechanisms
- Internalizing best business processes and management concepts, forming enterprise "internal strength"
At first glance, such an AI operating system seems extremely complex, but in fact, many basic modules can directly call existing AI tools, with the key being system-level integration and connectivity. The ultimate form of such products in the future will be deep evolution based on Feishu/Slack-like interaction experiences. The arrival of this paradigm will inevitably trigger profound changes in enterprise management thinking and productivity.