What Tech stack?

@twesolowski's post might interest you then: https://www.indiehackers.com/post/i-built-an-ai-startup-in-one-weekend-with-no-code-e766cbec13

If you’re non-tech do it with Bubble.
You can definitely build what you’re describing with it and leverage APIs to GPT

Bubble😀

Who's going to be building the app?

I recommend Laravel and React. Easy to scale, maintain and work with.

I have built multiple websites with it including PieSocket realtime, its always a joy working with Laravel!

I am using nextjs and Postgres for DB

Anything.

Nothing matters until you can give an MVP with as few features as possible to prove the hypothesis.

Once you've got a list of people willing to give you money, then I would focus on which tech stack to use.

For what it's worth, your FE/BE language/framework is irrelevant until you get to large scale. You should focus on what tech stack is easily hireable. I.E No point chasing for Rust Developers if no-one using Rust, and no point chasing for Svelte FE Devs if it's hard to find someone that knows it inside out.

Creating a journaling app with AI capabilities is a cool idea. The tech stack you choose can depend on factors like your team's expertise, scalability needs, and the specific AI functionalities you want. Here's a general recommendation:

Frontend:

React Native or Flutter for cross-platform mobile development.
React.js or Vue.js for web-based applications.
Backend:

Node.js or Django for backend services.
Express.js (Node.js) or Django REST framework (Django) for creating APIs.
Database:

MongoDB or Firebase for flexible document-based storage.
PostgreSQL or MySQL for relational databases if you need structured data handling.
AI & NLP:

Python with libraries like TensorFlow, PyTorch, or Keras for AI/ML capabilities.
For Natural Language Processing (NLP), spaCy or NLTK in Python can help with text analysis.

Google Cloud Natural Language API or IBM Watson for pre-built AI services if you prefer not to build models from scratch.
Cloud Services & Deployment:

AWS, Google Cloud Platform, or Microsoft Azure for cloud hosting.
Docker and Kubernetes for containerization and orchestration.

Remember, integrating AI features might require specific expertise or hiring AI specialists if you're not familiar with these technologies. Also, always consider user privacy and data security when implementing AI functionalities in your app.

You can interact with AI APis from basically any backend, so the real question is what programming languages are you comfortable with?

If you're comfortable with typescript or javascript, a Node.js backend and a Next.js frontend might be a good fit. If you're comfortable with dotnet, maybe give ASP Dotnet Core and Blazor a look.

As it relates to the AI APIs, I would definitely give the Microsoft for Startup Founders program a look. They give $2500 of free OpenAI credits to qualifying founders.

I gather that you're interested in utilising the OpenAI API or a similar service. You could start with no-code services, but in my experience, they're just temporary solutions.

If you plan on developing a full-fledged product, I recommend focusing on Python, Langchain, OpenAI API, and any database. This combination should suffice for all your needs, but having a mid-level developer on board is crucial.

in the end, the best tech stack to chose usually is the one you already know, because nowadays you can pretty much code the same thing with all the stacks.

But as a non tech guy, i guess that leaves you with few options: no code. Haven't tried myself but i guess you can have a look at bubble.

It depends on whom you can reach out to and have him partner with you - let him decide on the tech stack, you should care more about the non-technical aspects.

It depends of the "AI Feature" and the final propose of the app and how you think that it should be.

You can start with no-code platforms and when you have your customers and a MRR then you can hire specialists..

If you have a technical founder or friend to work with you then follow him. He will know.

For example, for a full based product Python stack with tools like pyTorch is a must. But for a just app that interacts with OpenAI or similar Javascript with React can do nearly 100% of the work.. It depends 100% of how you think the project and the underlying complexity.