3 ms·

I really liked http://getpopfile.org http://getpopfile.org . I had a somewhat related idea where I wanted to use ML to filter incoming emails and assign/tag the

by anujkk 14y ago

I really liked http://getpopfile.org http://getpopfile.org . I had a somewhat related idea where I wanted to use ML to filter incoming emails and assign/tag them as to-do list items. Have you considered making a web/mobile version? Email overload is one hot topic these days.

jgrahamc 14y ago
Actually what I considered doing was making it a service. Since it can all operate transparently through IMAP (including training and filtering) it's trivial to make it work with any IMAP based email. I have it working against my own gmail through OAUTH, but I decided that the market was likely too small.
- anujkk 14y ago
  How small is the market and how much work it will take to make a web/mobile app for it? I think if done properly it can get some paying customers. Even if it doesn't become a big startup, it has the possibility of becoming a good source of passive income. How about reimplementing the core logic in python and exposing it as library/API so that all kinds of web/mobile/plugin front ends can be built on top of it?
  - jgrahamc 14y ago
    I actually have the entire thing rewritten as a C library that I have been licensing commercially to companies for years. That part's easy. The hard part (for me at least) is building the web site, billing etc.
    - anujkk 14y ago
      May be you need a partner who is experienced and fluent in building modern web interfaces.
      - jgrahamc 14y ago
        If someone wants to license my polymail library for this purpose I would be happy to do that.
  - ryanlchan 14y ago
    It's small - I launched an MVP into this area recently. Business people who are inundated by email want a heavy, full-service solution like SaneBox, whereas consumers have been conditioned into thinking that everything regarding email should be free. There's not a lot of (paid) room in the market left.

3 ms·

I really liked http://getpopfile.org http://getpopfile.org . I had a somewhat related idea where I wanted to use ML to filter incoming emails and assign/tag the

by anujkk 14y ago

jgrahamc 14y ago
Actually what I considered doing was making it a service. Since it can all operate transparently through IMAP (including training and filtering) it's trivial to make it work with any IMAP based email. I have it working against my own gmail through OAUTH, but I decided that the market was likely too small.
- anujkk 14y ago
  How small is the market and how much work it will take to make a web/mobile app for it? I think if done properly it can get some paying customers. Even if it doesn't become a big startup, it has the possibility of becoming a good source of passive income. How about reimplementing the core logic in python and exposing it as library/API so that all kinds of web/mobile/plugin front ends can be built on top of it?
  - jgrahamc 14y ago
    I actually have the entire thing rewritten as a C library that I have been licensing commercially to companies for years. That part's easy. The hard part (for me at least) is building the web site, billing etc.
    - anujkk 14y ago
      May be you need a partner who is experienced and fluent in building modern web interfaces.
      - jgrahamc 14y ago
        If someone wants to license my polymail library for this purpose I would be happy to do that.
  - ryanlchan 14y ago
    It's small - I launched an MVP into this area recently. Business people who are inundated by email want a heavy, full-service solution like SaneBox, whereas consumers have been conditioned into thinking that everything regarding email should be free. There's not a lot of (paid) room in the market left.