The more interesting opportunities seem to be in solving industry-specific problems where AI can improve workflows, reduce operational costs, or automate repetitive decision-making.

Some areas that still look promising:

AI for healthcare operations and patient workflows
AI copilots for sales and CRM teams
AI agents for customer support automation
Predictive AI in logistics and supply chain
AI for real estate analysis and investment insights
AI-powered cybersecurity monitoring
Vertical AI SaaS for industries like legal, finance, or manufacturing

What’s becoming clear is that successful AI startup ideas are less about “using AI” and more about solving a real operational bottleneck with AI integrated into the workflow naturally.

It also feels like infrastructure, data quality, and distribution matter more now than just model capability alone.

Curious what sectors people still think are genuinely underbuilt in AI right now.