TuneR

Inspiration

Working with AI prompt platforms is great for quick brainstorming and conceptualization, but can be tricky when developing more robust MVPs that involve multiple architectural and decision plans. The larger the project and codebase, the more tokens might be consumed per prompt query. While enhancement features are helpful, the ability to not only receive a suggested prompt enhancement but also view predicted token consumption with optimization, a difference score and explanations extends usefulness of tools like Bolt's prompt enhancer. For context into inspiration for this project, see: https://discord.com/channels/364486390102097930/1387697063944257678/1387697063944257678

What it does

TuneR lets builders:

1. Input a potential prompt
2. Receive a suggested revision that can be copied and pasted as a prompt input
3. View explanations for the suggested revised prompt
4. View a comparison of the potential number of tokens consumed per prompt (original vs. suggested).
5. View a token usage analysis score to see the difference between the prompts.

How we built it

The project's current implementation was built with Bolt using React, HTML, CSS and deployment via Netlify.

Challenges we ran into

Tweaks and adjustments to the UI were made in a follow-up prompt to remove redundant features.

Accomplishments that we're proud of

I am happy to kickstart a purposeful and useful solution that builds upon an interesting concept for AI-native builders to leverage prompt and conversational-based engineering in modern app development.

What we learned

We're in the first wave of intriguing and useful tooling around prompt engineering. It is likely here to stay. Now the goal is to explore ways to leverage tools like Bolt for more robust AI-native application development from conceptualization, to mature MVP.

What's next for TuneR

Improve the UI and add functionality including:

1. Multimodal mode for input and output
2. A more robust algorithm for prompt enhancement, optimization, token usage and scorning.
3. Implement multi-platform capabilities (Bolt, Lovable, V0, BASE44, etc.)
4. Include multilingual views
5. Drag and drop features to include pre-sets and computer vision analysis for smarter multimodal functionality
6. Voice features for audible conversational prompting engineering
7. Import mode to automatically bring accepted suggested prompts into platform and project
8. Compare-and-Select feature to view suggested prompts, token forecasts, optimizations and select the best platform to use for the prompt and project with smart import to pull the project in and begin work.
9. More settings
10. Advanced preview, draft and publish capabilities
11. Model tuning for bring-your-own-model integrations
12. Connect to a database for proper auth, query and security to permit accounts and the ability to save projects
13. Opensource API

Built With

• bolt
• css
• engineering
• html
• lowcode
• netlify
• no-code
• prompt
• react