GraphSense – A Scalable Cryptocurrency Analytics Platform | Hacker News Reader

GraphSense – A Scalable Cryptocurrency Analytics Platform

graphsense.info

77 pointslainon10 comments

Looks interesting.

I would love to get involved on the design side. Just did some other work for a client creating a tool that allow you to analyze multiple cryptocurrencies and their underlying metrics. (Difficulty, close price, volatility etc) and are interested in exploring this field a little more.

Here is small section of one of the directions we went:

https://cl.ly/c997d4680e7c

https://cl.ly/a41354620e82

If you are looking for someone to help send me a PM (and yes for free I understand it's an open source project)

That looks nice. Did you use a UI/CSS library?

No all hand done

You should probably also check out https://cryvo.com/

I'm interested in Ethereum analysis because of the token economy. Specifically, I would like to see when a large amount of tokens are moved from a known exchange addresses.

Some friends of mine are doing this https://www.coinfi.com/signals

Thanks for posting that. It looks like a service traders would appreciate.

I'm looking for something open source though because my interest is intellectual and the "signals" type stuff is associated with pump and dump groups.

There is also so much market manipulation going on from all sides that it's really impossible to take any signals seriously.

Thanks for the feedback - cofounder of CoinFi here.

We actually help fund the open source Ethereum ETL scripts created by our data engineer Evgeny here: https://github.com/blockchain-etl/ethereum-etl

In addition, we've contributed the ingest as a public data set to Google Cloud if you want to do your own analytics without running the architecture: https://cloud.google.com/blog/products/data-analytics/ethere...

The compilation of exchange specific addresses is a long and tedious process, plus we hypothesize that exchanges would rather prefer to keep their addresses private, so we're not releasing that information though.

However, one of our Signals products reads from the Ethereum blockchain in real-time and notifies whenever there are abnormal movements to/from exchanges.

Unfortunately we're aware of the "pump and dump" association with signals - any suggestions for front footing that negative connotation and showing users that this is in fact different?

One way we've thought about is to create a Signals Library that basically explains each signal we provide and details the behind what it represents, a rough overview of what triggers it, and the historical performance if you traded off of it.

Any other suggestions would be greatly appreciated!

That's really fun! Funny enough, we put up a tutorial last week on looking at the silk road bitcoin / rogue DEA incident with Neo4j<>Graphistry (we do GPU viz): https://colab.research.google.com/drive/1TwYTDaBcMFxL6g5xLQz...

Would love to see if we can do even cleaner as GraphSense<>Graphistry!

Would love to play around with this, but getting this up and running seems daunting.

Reply on news.ycombinator.com