Had a good experience exporting sheets from my Tiller spreadsheet including an annual comparison sheet, net worth snapshot and transaction data to PDF and then uploading the as sources in Google’s NotebookLM. Makes it really easy to interrogate the data using AI. I’ve also tried this with ChatGPT but useful with NotebookLM not having to reupload source files each time you want to ask questions.
I’ve tried using Gemini with Tiller but it isn’t intuitive and any results take a lot of work.
Does show how powerful Tiller is with a very developed data layer, only a matter of time before all of the AI tools are integrated.
If anyone else has good advice on this front I’m all ears! Thanks!
@grifco Thanks for sharing the experience. I’m not familiar with Google’s NotebookLM. How are you “interrogate” the data? What are some of the features that standout? Is there any dashboard type functionality potential? Does “not having to reupload source files” mean once exported / uploaded that the data in the Tiller template is dynamically linked to NotebookLM?
The data is static once exported but exporting from Tiller –> PDF –> NotebookLM is a 60 second process.
After that you just type in questions like you would in Google but about your own data. So you could ask a general question like “tell me about my finances” or be hyper specific like “how much did I spend at a specific restaurant last year,” or “what are my key spending trends” - anything you want really.
Some AI platforms have a library functionality that allows you to upload files to work with, that is what NotebookLM allows. Then if you wanted to ask questions and have the AI answer questions with a certain philosophy for financial planning you can also upload sources files for it to draw from (e.g. whatever financial planning books you like if accessible in PDF).
This is undoubtedly what the future of Tiller looks like but we will have to wait a little while for that. In the meantime this is a good workaround.
