LeisureAI: AI powered personalized calendar

[maeve]

Name of Project:
LeisureAI (LAI) : MVP Phase I

Proposal in one sentence
LeisureAI is an AI tool that helps busy people and/or digital citizens find IRL events and interests that enrich their lives.

Description of the project and what the problem it is solving:
With so much happening around us at lightening speed, its difficult to step away from our digital lives and enjoy engaging with our hobbies and interests IRL. LAI fixes that problem by scanning and aggregating from event sites and platforms for relevant interests that you determine while interacting with it.

Grant deliverables:
Define user personas and product workflow
Finalize requirements
Hire assistance to design a prototype

Spread the Love:
I am seeking a data scientist, ML expert and blockchain architect to help me define the framework and technologies to build a prototype.

Discord handle: maeve#4318

[metamyth]

Hey, cool idea @maeve

What you’re trying to build is a recommendation engine with a calendar UI. Graph neural nets and structures encapsulate social interactions very well, I suspect they can also be tuned for people-events dynamic as well.

This product also precedes something like matching in social pools at the events model recommends.
For the project, I think you’ll need some ML/DS folk who can sort out the processing when you have data with your targeting.

I’d be happy to see this go through!
metamyth

[orpheus]

WIll we be able to follow the progress on Github? I’m personally interested in how the events-data extraction is performed, from (I assume) the dataset of events websites.

[swapneelm]

One thing that would help me better understand why there’s a blockchain architect required would be a more detailed explanation of how you intend to track user interests and the advantage of using the blockchain to develop a recommendation system (assuming that’s the key novelty?) versus scraping a good old tabular dataset and using an off-the-shelf rec-sys. Also what would this system improve compared to using existing platforms like meetup.com (or figuring out a real-time alerts / rec-sys on top of users’ existing meetup accounts).

Another useful link to add in would be what datasets you’re looking into creating because that by itself sounds like an expensive task given that you’ll want to do it in real-time as events pop up. So the computational setup might also be a bit expensive to set up.

Re: aggregation, using APIs might be a good bet; here’s an example: meetup-dataset/README.md at master · moxious/meetup-dataset · GitHub

Overall, really good idea – but would benefit from a bit more early scope definition and sharing of links. I do understand that from the current description (“MVP Phase I”) the grant is meant to give you the time to figure out the prototype, but at least an early version of that is something you want to do before applying, to convince people you are capable of delivering. Happy to discuss

[maeve]

Hi metamyth:

Thank you for your feedback, this is an idea that I’ve been excited about for several years now, and I’m glad to have a chance to work on it. Definitely interested in connecting w more ML/DS people to make this a reality so we don’t have to do manual searches to broaden our horizons {not to mention getting out of the house:}

Thx so much for your support! :}

[maeve]

Hi orpheus:

Yes the current plan is for this to be open source, however that’s also subject to design viability once I onboard technical talent. I’ll certainly keep the community posted on progress of this if it gets funded.

Thx for your question! :}

[maeve]

Hello swapneelm:

Great feedback, thx! A blockchain architect is not necessary to begin, however I do want to explore potential integrations that could enable many more use cases that I foresee in future iterations and since Blockchain IS the future… This is different than meetup (I had this idea long ago, before I knew meetup existed), the difference is that users won’t have to take the time to go seeking information on variety of platforms, instead information comes to the user in a format that helps to organize their edification, growth and desired lifestyle. The user won’t have to know where and how to find information, they just need to define parameters and interact w their “helper”. Also, it could be about meeting people w similar interests, but it doesn’t have to be.

I’m very capable of delivering a prototype w the right team. :} problem solving is my super power :} Yes pls reach out - would love to discuss more!

Thank you again for your support!

[john-mac.eth]

Nice idea @maeve

An AI app that can sort these events by user preferences or say “wish list” would be incredibly useful. I like the idea of filtering events by relevance to these preferences, rather than sorting through voluminous invitation emails to find events that meet interests.

As @swapneelm mentioned, an API call for data aggregation seems like the best solution based on simplicity and low cost.

Well done!

[orpheus]

Feedback on the name: In my case I would like to track ‘scholarly’ events (reading groups, conferences, etc) and so the name wouldn’t make sense for that.

[maeve]

Hi john-mac.eth:

Thank you for your comments! :}

[maeve]

Hi orpheus:

The original name was much longer :} Leisure implies whatever you want to do in your spare time, reading groups and conferences included. It can be used for professional services, work resources or whatever you like… who am i to judge! :}