• Community
  • Interviews
  • Podcast
  • More
Community Talk shop with other indie hackers.
Meetups Meet indie hackers across the globe.
Products See what everyone's working on.
Interviews Learn from transparent startup stories.
Podcast Listen to raw conversations with founders.
Articles Read tips and advice from your peers.
Contribute Share your knowledge and experiences.
October 9, 2018

How I built Defects AI, a machine learning service to predict GitHub labels.

11
votes

How I build Defects AI a machine learning service to predict GitHub labels. – Indie Hackers

Defects AI is a service to predict labels for a GitHub Issues fully automated, with machine learning models trained in particular to that use case. Don’t

Sascha Heyer

4 replies


  1. 3

    The posted link https://www.defects.ai/features/feature/automatic-issue-labeling doesn’t seem to be working.

    localheinz
    ·
    16 hours ago
    ·
    reply
    1. 1

      Hi Andreas,

      confirmed :) fixed. Thank you.

      Sascha Heyer
      ·
      16 hours ago
      ·
      reply
  2. 2

    Looks awesome! I did a quick test worked like a charm. Best of luck 🚀

    depomoty
    ·
    16 hours ago
    ·
    reply
  3. 1

    Cool! You could extend this by implementing a seq2seq model that summarizes github issues and also recommends related issues:

    https://towardsdatascience.com/how-to-create-data-products-that-are-magical-using-sequence-to-sequence-models-703f86a231f8

    philgooch
    ·
    3 hours ago
    ·
    reply
by @csallen at Stripe

Learn from the founders behind hundreds of profitable online businesses, and connect with others who are starting and growing their own companies.

© Stripe · FAQ · Terms · Privacy ·

Community

Meetups Top Discussions This Week This Month Join

Products

All Products Highest Revenue Side Projects Self-Funded Add Yours

Learn

Interviews Podcast User-Submitted Contribute