Not sure if this is an established hobby or something I've just come up with myself, but I've been "dashboarding." Essentially I have a Blazor webapp that integrates lots of data sources (some manual, some automatic) from areas of my life and I use that to visualize and analyze goals and habits. The main page consists of rolling-weekly stats that deliver "integration scores." Each score contributes to an overall score that gives me a general idea of how I'm doing on all my habits and goals.
So for instance, I use YNAB for our family budgeting, and I have it setup so that if I go a whole week without performing reconciliations, I get dinged -1. Otherwise this sits at 1.0. Then I have a score for journaling - my goal is to journal 4-5 times per week, so each time I journal it resets the score to 1, and then slowly ticks down to 0 over time. Then I have a number of Apple health scores that get imported automatically via REST API. This part compiles all the data on calories, relevant macronutrients (I mostly track protein and fiber currently), steps, workouts, etc. and builds a nice visualization. I consider a total integration score of 0.8 to be pretty good - keeping at that level is actually better than seeking for a perfect 1.0 all the time as my theory is that it will prevent burnout and allow for some forgiveness, because I can't be perfect.
It's been a fun project, and one that I generally try to avoid any AI use. Fun to just build and because the stakes are so low I just chip away at one feature at a time, carving out 15 minutes here or there.
Do you have more details on this project anywhere? I've been working on habit-building and tracking in my journal for the past year and a half or so, but I'm looking to amp it up a bit more. Your project appeals to my software developer and hobby collector mindset and would love to learn more about it.
Not really, but I'm happy to share more here :)
In all honesty, under the hood it's a bit of a mess. I may have eschewed some of the software engineering best practices in lieu of building something quickly that I wanted. I'll get around to going back through and retrofitting the app with some cleaner code, but for now I couldn't even open-source it without a self-perceived hit to my portfolio.
The project largely started out as something else. I initially wanted a combined TO-DO list and journal. Rather than checking things off I would run the journal content through a local LLM and have it check things off for me based on what I wrote each day. That's yet to be implemented. Then I moved on to an "ordering" system - I was inspired by the way that medical practitioners put in orders once they determined a course of treatment, and thought that might be a useful model to help motivate me to get things on my list done more effectively. I built this, but have utilized it less than I thought. Since then it's mostly been focus on the integrations and scoring system. The whole thing is highly modular, so for each integration I grab a template for the visualization I want to build and then need to reason out how to get the data into the system, which usually involves an API integration, scraping from some online data source, and/or data engineering. It's very fun, because each integration module has its own challenges.
I built the app using a standard stack of .NET core, Blazor server, and the data is stored in SQL server and data operations are handled with EF core. I use the Radzen component library, which I like a lot from a developer perspective but it's challenging to retheme and I'm largely unhappy with the look/feel of the app. This is something I plan on getting to eventually.
Happy to answer any/all questions. It's such a personal, homebrewed app that I can't imagine anyone else would get as much use out of, but it's very powerful and I think the hobby aspect of it could translate to pretty much any other developer.