I’ve been working on an AI agent called Splitly, designed to help people manage personal finances directly through chat. Splitly can record spending, bills, and borrowed or lent money. Instead of just typing, users can also upload bill photos or send audio messages, and the agent extracts and logs the details automatically. It then provides summaries, balance sheets, and recommendations such as highlighting overspending, or suggesting limits.

Splitly lives inside NexChat, a messaging app we’re building where people can connect not only with other users but also with AI agents. Users can interact with Splitly privately as their personal finance assistant, or they can invite it into a one-to-one chat so that two people can collaborate with the agent in real time. Beyond Splitly, anyone will be able to build and submit agents to NexChat, making it an open ecosystem where agents can become active participants in conversations.

The broader goal is to treat AI agents as first-class chat participants, useful not just for individual productivity but also in shared, human-to-human contexts.

I’d appreciate feedback from the HN community on both Splitly and the idea of an open agent ecosystem in chat.