Ask HN: Deep Learning startups made by indie founders? | Hacker News Reader

Ask HN: Deep Learning startups made by indie founders?

8 pointsrayalez11 comments

Hey everyone! I'm interested in DL, but it's hard to come up with practical applications for it that I can create on my own.

It would be really cool to make a project that utilizes the recent advances in DL(seems like a big opportunity), but all the applications that come to mind seem to require a ton of resources(data, compute, scientists, developers), and are useful only for huge companies(self driving cars, search, spam, face recognition, etc).

Are there any DL-centered startups made by solo founders on small budgets?

What are some ideas or areas where a solo developer can make something useful and profitable?

View on HN

Here is one idea: stock trading robots - no customers, no employees... Just great location independent lifestyle business and the brier to start is pretty low: linux box + historical data...

I have been doing it for years and I am earning much more on this "side gig" compared to my "real job" or maybe I should call my full time job a "side gig" at this point...

where to get historical data?

where to get real-time data?

what api allows you to trade stocks?

Real time data here are two that are free:

https://www.alphavantage.co/documentation/

https://iextrading.com/developer/docs/#getting-started

The majority of discount stock brokers provide API, here is one example:

https://www.interactivebrokers.com/en/index.php?f=5041

Historical data, the cheapest I can find 20+ years of data fro less than a $100 was on eBay, just search for: "us historical stock market data"

Thank you!

Depending on your needs it's possible to leverage Robinhoods API for trades. Might be against their tos tho

Would you do mentoring? I am interested in doing it but have NO knowledge whatsoever in stock trading. Currently am part-time final year postgrad of machine learning and data science.

edit:spelling error

I am sorry, but I cant and you should not look for this either... Here are the 2 major problems with mentorship:

1. Your mentor will never tell you a working trading strategy, instead he is going to use for himself to make money... You are his competitor in the stock market, so he'll feed you BS to get your money one way or another.

2. If you end up discovering a good trading strategy and share with your mentor, he'll steal from you, discourage you to trade it or even trade against you...

Sorry to say, but there are no short cuts and you'll have to put the work to reap the benefits... There are no friends in the stock market..

All are good points. Thank you.

Https://SignalBox.ai

My deep learning startup, self funded, I wrote all the software myself

Recently just built and launched hydra too hydrachain.io

How is your startup doing? Can you share your story? Maybe post in IndieHackers?

I am interested in freelancing/consulting in data science and deep learning but don't know where to start.

it was a slow-ish 6 months, but I'm cash flow positive, but in the last 2 months after the launch of my second product it's gone crazy - both Hydra and Signalbox

this month I've 4x revenue, next month is 20x already booked, I turned down an acquisition offer 2 weeks ago from one of our customers, and next week I had to cancel speaking at an AI pitch conference thingie because I'm going to discuss acquisition from a different customer, wish me luck

I would like to tell my story, but frankly, my todo list is mental. I've just been interviewed by the telegraph, a UK newspaper, so there'll be an article there

https://ramm.science/ for those interested

Reply on news.ycombinator.com