What are you building in AI

Hello dear community

We live very exciting times at the moment and there are tons of opportunities out there . What are you building in AI?

An AI tool that let’s you communicate with the documents

We are actually raising $1.5m from a Bubble prototype that uses A.I. to increase the efficiencies and utilization of surgical units and surgical centers. Modules include staffing, scheduling, inventory, billing/claims and more all connecting data to A.I. for real time feedback of decisions to be made by all of the various end users at once. We started testing it about 17 months ago.

Final version will probably not be built in Bubble, however, without Bubble we would not have been able to test the model and proof of concept.

3 Likes

I’ve created a prototype alpha release for a product that mirrors key functionalities of bubble to allow a user to input their issues and it will troubleshoot or generate a product for the user.

Functionalities are currently a UI color schema generator with example UI, Database generator based on type of app you’re building (with entity relationship diagram), and a copy of workflows to troubleshoot bubble workflow.

Built bootstrapped. No external funding. Definitely seeking feedback

2 Likes

I know about one pdf.ai Very interesting

This is amazing @underhill.dan Congrats!

1 Like

https://flexgpt.io built on Bubble with the ability to chat with docs :slight_smile:

2 Likes

@georgecollier I guess the website is still in the development phase. Seems like a nice solution

I am building an AI business consultant that empowers SaaS start-ups to grow and succeed by providing them with actionable insights and strategies.

Started building a couple weeks ago and I have built the first of many features that helps users select a target market for there SaaS, all you have to do is answer a few questions about your SaaS idea and it generates a target market for your SaaS as well as giving actionable steps to validate if this target market is the right fit for your SaaS.

After this initial response from the AI the user is free to interact with the AI to further refine the response.

If anyone wants to see a demo or hear more shoot me a message or an email at bodie@saasight.com

why not drop a link here?

We’re working to create tools that can measure aspects of people’s spiritual practice, wisdom, and trustworthiness, and then help provide them content that can increase their progress along the spiritual path.

2 Likes

It’s great to see that many AI apps are being built using Bubble. Recently, we helped a client launch their AI web app MVP (which focuses on semantic search) using Bubble. We also have a Chrome extension in the pipeline to save and explore 1000+ curated prompts, which we are planning to launch in two weeks. If anyone needs expertise in building AI apps with Bubble or using code, feel free to reach out to me. We’ll find a way to help you out.

@accounting4 Very inspiring. Could you please tell me a lil more about it?

The title of this thread says all with the current hype - the solution/technology is discussed before defining the problem to solve.
“AI” (or rather ML) is not a silver bullet and won’t do magic.

The real questions to ask are :

  1. What’s the problem to solve?
  2. Do we have a large data set with validated data / or a model already available?

If so, ML may be considered.
However, too often, ML is used in-lieu of a good ol’ algorithm with traditional programming.
In cases where ML accuracy is less than traditional programming, then better off staying away from it, you will save a lot of time and effort.

I have some of my clients going back to traditional algorithm after “AI” failed to classify accurately a large dataset whereas a simple search * contains <value> would have sufficed.

From my perpective in its current form, ML must be used to compare/generate data against/from a large data set.

Everything else, until real AI comes, will vanish.

1 Like

This isn’t necessarily true. The nature of LLMs like GPT-4 is such that a dataset is no longer required.

For ML activities like prediction, classification etc, you do need a dataset, but prediction/classification is only a part of AI. Generative AI rarely requires a huge dataset (and a huge dataset can make it more challenging due to context limits)…

Then just use that. Nobody’s saying to use AI for everything. If your clients AI and you didn’t just suggest that a simple keyword search would suffice then that’s on you :stuck_out_tongue: Only recommend AI if it’s really necessary.

We are both correct.

If you don’t have a large dataset, then you may reuse a model which has already been trained on another dataset.
I have amended by answer to include model reuse.

Of course it was before I stepped in :grinning:.

2 Likes