Ask HN: Looking for ideas... | Hacker News Reader

Ask HN: Looking for ideas...

2 pointsmhlakhani8 comments

I've always wanted to build something interesting (web apps are okay, but something more towards research/number crunching/cloud computing would be better), but ideas are always a stumbling block. I just can't seem to find some good idea to work on. Usually what ideas I do get have already been done by someone else.

What would you guys do in this situation? Where do you get your ideas from?

View on HN

A few ideas:

* build a easy and very powerful ETL tool for data cleansing

* take genetic data (e.g. from 23andme) and offer people information on work related characteristics (like leadership ability)

* offer a service which predicts traffic jams and estimate how long it will last

Anyway, look for an industry sector which interests you. Learn about it (reading, talking with people, etc) and you will probably see problems which could be solved.

Thanks for the tips, but as a student it seems a bit hard to get access to the data sets etc. needed for those. However, I'll look into it.

You're welcome! I think I see your problem in 3, (2 isn't too hard, you can buy your DNA sequence for ~$200). There are a lot of logistics companies which may have valuable data but can't mine them properly (especially the smaller, not Fedex or DHL).

I don't know which industries interests you, but there are some industry sectors like logistics, agriculture or finance which heavily rely on new insights from their data.

You could start as a consultant for one company and build a product (consultingware). If it works successfully offer it other companies (http://nukemanbill.blogspot.com/2008/06/how-to-sell-your-sof...).

The blog post was a pretty good read, thanks. I'm interested in data mining/AI, but need to learn much more before I'll be able to do anything useful. After some more thought I think it might just be better to focus a bit on academics and learning the theory behind all this stuff first.

Partner up with someone like me :)

At any given time I have between 3-4 ideas that are ready for some type of development, and probably another 2-3 that need some refining before they would be worth considering for development.

If you want something to work on (of course if it interests you), shoot me a message and we can work something out.

Thanks for the offer, I'll be emailing soon

Watch out for leanwebsites.com launching by Feb 22nd to address the "ideas" problem.

I'll definitely keep a lookout, thanks.

Reply on news.ycombinator.com