Anyone else struggling with data in agent workflows?

[verodatai]

Hey everyone!

We’re doing some research into how people are using data when building AI agents—whether you're using AgentGPT or integrating other tools like LangChain, CrewAI, or AutoGPT.

A few things we’re curious to hear more about from the community:

What kind of data are your agents using today?

Are there challenges with accessing, structuring, or updating that data?

Have you come up with any clever solutions or workarounds?

We're exploring these challenges at Verodat and created a short, 1-minute anonymous survey to help us gather broader input from builders:
👉 https://verodat.typeform.com/AgentSurvey

We’ll post a summary of the insights here once responses are in. Would really appreciate your input—and hearing what you're running into (or solving!) on this front.