Bitcoin Forum
October 28, 2024, 04:19:31 PM *
News: Latest Bitcoin Core release: 28.0 [Torrent]
 
  Home Help Search Login Register More  
  Show Posts
Pages: [1]
1  Alternate cryptocurrencies / Announcements (Altcoins) / [ANN] Bonobot - Cutting edge machine learning bot on: December 25, 2019, 05:12:41 PM
Foreword
A couple months ago I tried (and miserably failed) to crowdfund my bot.
I'm now really close to a V1. Before official launch I'm looking for collaborators.

  • Designer (Ideally web dev, with skills in web design, but I can handle the dev part) Paid (well, a share of the license income)
  • Community manager(s) (Ideally at least one native English speaker, I'm (obviously?) not) In exchange of free licenses.
  • Moderators In exchange of free licenses.

You can shoot me a mail: potatoe.crowdbot@gmail.com

This very announcement is aimed to be reworked, with a neat logo and better English.

What is Bonobot ?

I spent one year trying and implement state of the art techniques of unrelated domains and apply it to trading.
I now confidently say that I'm on something hugely powerful.

To give you a gross idea, here's what the backtest on test data (ie data that the models have not seen during the training phase), gives on ETHUSD:
 

So... yeah...

The trading part is usable, I'm currently finishing the GUI, adding charts, statistics, advanced reports, backtests...

Side note: Since there is few old school algo and a lot of neat AI, contrary to other bots on the market you won't have (at least for now) a lot of handle on the config parameters. For now it's limited to: bet size factor and pair (+Obvioulsy API keys of the exchange platform)

When ?

Official first release needs a logo, a website, a licensing system. Depending of the help I get on that, it can take a few days to a month, so realistically in the course of January.

How much will it cost ?

In its current state, it's not a get rich quick thing (I wouldn't be selling it if things were that simple ^^).
That said I truly believe we'll get there.
If too much people use the same best algo in the world... well... it won't be so for long.
For that reason I'll only release a very limited number of licenses, plus a limit per pair and platform.

Also, I'm confident enough to put in place a refund policy (It has to be thoroughly thought and carefully crafted. I'm working on that with a lawyer friend of mine).

The offer would go as such: $50/month for 1 pair on 1 platform + $10 per additional pair/platform.
It will give you access to
  • Updates (improvements on a weekly basis + bugfixes whenever needed)
  • Access to private Discord

It goes without saying no one will have to pay anything before everything (Usable v1, Discord, website) is in place and working !

You're in ? send me an email at potatoe.crowdbot@gmail.com with the subject Bonobot FTW, tell me the exchange platform you'd use and the pair you want to trade.
Seats are in limited amount, first-come first-served
2  Economy / Trading Discussion / Crowdfunded cutting edge machine learning bot on: August 18, 2019, 09:01:48 AM
Here's the full story: https://bitcointalk.org/index.php?topic=5176445.0

But it's long, so here's a TL;DR version:

  • I'm a coder that wants really bad to be a machine learning researcher.
  • My trading bot is in the state "Very promising proof of concept".
  • I want to crowdfund a whole year studying and apply learned knowledge 100% to the bot.
    Contributors get to use the bot (+ constant update).
  • Crowdfunding would allow a meticulous approach and pressure free clean code, so, the first 3 months would be just updates (+ a version of the bot you should really only use with paper trading).
  • ~3 months mark: usable version.
  • ~6 months for a beginning of maturity.
  • then 6 months improving, aiming for awesomeness.

Original post (link above) will give you more details.
If you have any questions, you can ask it in this thread, via DM, or even mail (potatoe.crowdbot@gmail.com).

Seats are limited (I want to eat and pay my rent, not make a business out of this) and contribution will be in the range $30 ~ $50 (USD)
 
It's here to be part of the adventure: https://forms.gle/PMUmoHyqi6yxAunA6 



(Because you were kind enough to read me, here's a hacker trick you can use to impress your colleagues and friends (well... it's just a gmail feature :p but I bet a lot of you don't know it)


Here my email address: potatoe.crowdbot@gmail.com (I created it for this project, but I guess you figured that out)
And here's my email address p.otatoecrowdbot@gmail.com, and p.o.t.atoe.crow.db.ot@gmail.com , oh, and potatoe.crowdb.ot+ILovePotates@gmail.com... whatever  you're gmail address is, you can put a dot between any two letters, only letters count and it will be received my the same gmail box. Also you can insert the character + followed by anything you want, it will be received.
You're a hacker too now  Grin welcome to the club.




PS (to the intention of moderators): I was note sure to be posting in the right sub forum in the first place. I hope the content will prevent this post to be considered as a spam. ie I'm crowdfunding studies to become a machine learning researcher. While I totally get someone trying to promote a product for a company in the very same way would be flagged as a spammer, I hope I won't be, and I'll totally get it and won't complain if you decide I am Smiley
3  Alternate cryptocurrencies / Announcements (Altcoins) / Crowd bot. Crowdfunding my career change+Algotrading on: August 17, 2019, 03:44:02 PM
(English is not my native language. Though I really have no idea how obvious it is.)

I'm an experienced coder (I began as a web dev, then went to software architecture. I also worked as a security consultant).

Since about a year, I got increasingly interested in machine learning.
After a dozen of books on the topic, hours on Coursera, I achieved, I believe a pretty decent level.

Further from absolute mastery that I'd like though. My objective would be to master the field enough to do some research (on evolutionary algorithm, that's kind of my actual life goal)

To give you a rough idea of Where I'm at, I'm able to read research papers and more often I would be able to implement what they talk about from scratch. (Which would require a sick level if Google was not a thing, yet still not the simplest thing on earth with it)

Now, here's a question:
I want to spend a whole year learning like crazy. Learn, and apply all acquired knowledge to one and only one project: Algorithmic trading. What could that give ?

The answer to that question is probably pretty straightforward, and goes along the lines of:
"Dude, I don't even know you, what fuck should I give ? You moron."

That's rude, character I just invented for the sake of my argument, but totally fair.

Now, what if I told you, you get to enjoy the fruits of my work ?

People crowdfunding their studies is apparently a thing. To be honest, I'm not sure I would be kind enough to help a stranger in that way without any counterpart.

So here's a counterpart: I try to code the best trading bot ever created, and you can use it.

(Which doesn't tell you much, this is probably the case for anyone coding a bot. I mean, no one is trying hard to be the third best.)

That, I assume, raise some questions. Here are the ones that I would have wanted to ask, should I've been you reading those lines:
Will you succeed ?:

It depends on the goal.

I've we're talking about the best bot ever... That's... quite unlikely.

If the goal is to make every contributor much more money that they put in. I'm actually very confident that will be the case (not without reason, more on that latter), if we're talking about a large amount of money, that's a real possibility, but that's my opinion. Which leads me to the next question:  

Why should you trust you ?

That's a really good question, Here's a two parts answer:

1) You shouldn't

Let's be real, from your point of view, I'm just a dude on the internet.

"Just a dude on the internet" is a category of people among the least trustable.

When it comes to trust, "just a dude on the internet" ties with "Convicted pedophile applying for a babysitter position".

But there is even worse: "Just a dude on the internet who wants to make you rich".


2) On the other hand


I asked myself what would be susceptible to raise my trust at your place, and here's what I come up with:
  • The funding will be monthly (Patreon type of thing), which is probably obvious yet it feels better making this explicit
  • We'll have a mean (either slack or discord) to interact directly, and I intent to give frequent updates
  • A web page showing real time results of the trainings will be online 24/7
  • I will, one way or another, make my identity verifiable. So you're sure I'm not from a far country you can't even pronounce (Well, unless you can't pronounce "France")
  • The number of seat will be limited. Which is not actually about honesty, at this point, I absolutely have no idea if that thing could work at all. But in the unlikely case this is a huge success and a lot of people are willing to contribute. A large amount of monthly cash will look like an unbearable pressure to me. Consequently the upper limit will be about what I need to my exact current lifestyle (which consists in paying my rent and buying the food I want)


I won't take a dime before those conditions are met

Is this a get rich quick thing ?

I wouldn't need you if it was. In fact, I want to take my time to implement everything cleanly, with tests, trades reports, a decent UI.


At this point, there is a detail that seem of importance. I told you I was confident for reasons: here why: the bot is very advanced already. I even already made some profit. I also told you I wanted to do research (point on which I insist, this is pretty much the goal of all of this). Well, some of the methods used are implemented from scratch and are my own creation.
It goes from super accurate to weirdly unpredictable. The whole code is drafty and needs to be cleaned, some part rewritten, I need to read a lot of research papers and math books. It "kind of" works, but I need to settle down, and be able to focus on writing great code without the pressure of having to make money immediately. And that's where I could use your help

The timeline would go as such: IF there is enough people to support the project, I will need a week or two to handle the requirement I talked about.

Then it would start, and the first three months would be the more boring (for you, for me that would be intense Smiley ).

Expect at the end of this period, a first usable version, three more months until it's decent, then six months to works my way through awesomeness. That's the plan anyway.

Three bits of information to go:
How much would it cost to be a contributor ?: I'm not sure yet, somewhere between 30 and 50 USD.

It's over here: https://forms.gle/PMUmoHyqi6yxAunA6 you just have to give me your email. And if you are not sure, or very sure, or absolutely whatever, there is a field for comments.

If ever that have any success, number of seat will be very limited. First-come-first-served
Pages: [1]
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!