Improving Lumo can focus on making it faster, more accurate, and easier to use in daily tasks. Better understanding of user questions, improved context memory, and more natural responses would make it much more helpful. Performance optimization is also important so it works smoothly even on slower devices.

It would also be useful if Lumo could connect with real-world services and provide practical information when needed. For example, integrating with transport or travel systems like Saudi Railway Company could help users get more useful, real-time assistance.

Overall, the goal should be to make Lumo simple, smart, and reliable for everyday users.
https://saudirailway.net/

You can now prompt Lumo AI through the https://lumo.proton.me/?q=%s scheme.

This would be awesome. I've actually been enjoying the ease of being able to use perplexity with just a @Perplexity in my search bar. If lumo could do the same thing, then I'd feel far more compelled to switch models.

The current process to prompt lumo is as follows:

1. Type lumo.proton.me
2. Page loads
3. Prompt Lumo
4. Lumo generates response

For simple prompts, a smoother way of interacting with lumo would be to...

1. Type keyword and prompt
2. Page loads and immediately start generating response to prompt

Example of how it could work:

- https://lumo.proton.me/search?q=%s
- %s is replaced with the prompt

Confused? This is a video of the feature being used in Perplexity.
https://www.youtube.com/watch?v=JTwAbD5OCBM