A New Approach to Enterprise Software

A Kenyan startup called Lua is challenging the traditional structure of enterprise software with a novel approach centered around AI agents. The company recently raised $5.8 million in seed funding, led by Norrsken22 and including investments from Y Combinator and Flourish Ventures.

From Assistance to Execution

While most enterprise AI today focuses on assisting employees—through chatbots or productivity tools—Lua is building software that performs work autonomously. Their platform allows organizations to create agents that handle complete business processes, such as customer onboarding or loan processing, without human intervention at each step.

This represents a shift from the traditional approach where complex tasks are broken down into smaller components handled by different systems and people. Even with existing automation, workflows often remain fragmented across departments.

How It Works

The Lua platform enables companies to build agents that can:

  • Take end-to-end responsibility for tasks like loan applications or insurance claims
  • Collect data, apply rules, and make decisions autonomously
  • Operate through familiar channels like Slack, WhatsApp, and email
  • Escalate only when facing uncertainty or complex situations

Early Adoption in Financial Services

Early deployments in Kenya have focused on financial services where manual processing delays are common—some banks take 3-5 days to process simple retail loans due to KYC checks and document verification.

As co-founder Lorcan O’Cathain explains, “We’re in the race to shape how human-agent collaboration gets defined globally.” He envisions a future where organizations consist of blended teams of humans and AI agents working together seamlessly.

Reshaping Corporate Structures

The rise of AI agents may blur the lines between technical and non-technical roles, as employees increasingly manage and improve these autonomous systems rather than performing individual tasks. Some companies are already seeing one or two people effectively supervising dozens of agents.