Visual Reporting Bot

Making reports while in the field can be complex for construction workers and other field agents. We built a simple experiment demostrating an AI Agent helping a construction worker report issues through Telegram, via image and voice memo.

Description

Combining generative AI and no-code is increasing the % of automation projects that can be ROI positive, especially for SMBs.

Take “reporting issues via a mobile phone”, for example. If you have a small team, you’re probably solving this problem via communication tools like WhatsApp, Telegram or Email (e.g. “I’ve sent you a few pics earlier today!”).

Here’s a workflow we’ve prototyped that could multiply your team’s productivity solving this task:

• User takes picture
• User records voice memo
• LLM (Large Language Model) merges both inputs and automatically creates and categorizes the issue

In the demo below, you can see how this would work in a hypothetical scenario where construction workers use a simple Telegram Messenger agent to report any issues they find.

After using the relevant LLM with a custom prompt, issues are automatically created as records in an Airtable database. These could also be created directly in your project management tool via a simple API call.

Here’s a quick video illustrating the process from the user perspective:

Potential

The workflow can be adjusted to multiple use cases: from reporting issues found in a physical location (factory, construction site, urban area….) to helping workers complete inventory related tasks (warehouses, retail stores, hospitals…).

In this particular example, the agent is available via a Telegram, but the functionality could be integrated in a web or mobile application.

📖 Original Article published on LinkedIn

👉 Need help improving your employees’ productivity? Schedule an intro call to discuss if we can help you too.