0zAND1z

@0zand1z

97 Following

16 Followers

0zAND1z pfp

"son, we got rick rolled in the ai generation too. no one must be spared." - sometime in 2047

0 reply

0 recast

1 reaction

0xdesigner pfp

hey sorry i didn’t respond to your text i’ve been trying to process my decision to commit my life’s work to an industry that has only found product market fit as a casino that requires a phd level understanding of computer science to use

25 replies

12 recasts

123 reactions

0zAND1z pfp

DAY 12: bm, fellas! 🔵 i resumed labelling the database today and came up with the first iteration of label. 💿 here is jesse's sample label on ipfs with proofs designed for readability and interpretation: https://bafybeifkibgoxfs5s3ox4pvakyrz7aazpyyftlelp2xxcl3ojubxgkzmq4.ipfs.w3s.link/sampleflair.json EOF 1/n

0 reply

0 recast

1 reaction

0zAND1z pfp

DAY 11: based sunrise, builders! 🔵 today i was busy working on labelling the dataset. that's it. that's the update. nothing fancy. 😄 will explain in detail once i finish labelling them. EOF 1/n

0 reply

0 recast

3 reactions

0zAND1z pfp

this is some win xp and qbasic ptsd stuff here. love it!

0 reply

0 recast

0 reaction

0zAND1z pfp

OᑎᑕᕼᗩIᑎ & onwards. 😎 to da sun, dis summa! 🚀🌞 EOF 4/n

0 reply

0 recast

0 reaction

0zAND1z pfp

and today, i spent time designing a k-means clustering mechanism to group the wallet's behaviour into natural group formations. excited to see some *psychology x web3* here! more on this will be posted as i make some more visible progress. that's it for now. kinda feel exhausted, but i remain motivated :) 🦾 3/n

1 reply

0 recast

1 reaction

0zAND1z pfp

i also optimised the dataset and started working on labelling the dataset based on the first version of label and qualifier criteria. here is a quick sneak peek of the criteria we will use to label the dataset: 👀 ps: the same labels are used in the graph (above) for visualising the wallet's reputation. 2/n

1 reply

0 recast

0 reaction

0zAND1z pfp

DAY 8, 9, 10: gm after a long based weekend! 🔵 i've not been feeling great past few days but here's a consolidated update from Saturday to Monday: i am exploring new ways to make the reputation graph look a bit more fun and align to web3 expectations. here is a revised "pixelated" version for your feedback: 1/n

1 reply

0 recast

1 reaction

0zAND1z pfp

possibilities ✨🧙🏿‍♂️

0 reply

0 recast

1 reaction

0zAND1z pfp

i've also enhanced dataset services, started mining txn data, and updated the backend api services to deliver sample reputation graph back to the client. ✨ hope to optimise the datasets, make the graphs look cleaner and continue building the pipeline over the weekend. a fun weekend grind is ahead of us! 🌈 2/n

0 reply

0 recast

1 reaction

0zAND1z pfp

DAY 7: bm, fellow buildooors! 🔵 i took up the eth:external labels today and started working on a sample reputation graph. each vertex is essentially a color-coded label! 😊 its very early as you can see, so pls do share your feedback: (ps: vertex color matching gradient edges are on their way) 🛠️ 1/n

1 reply

0 recast

2 reactions

0zAND1z pfp

0 reply

0 recast

1 reaction

0zAND1z pfp

wen /cryptoBLR ? 👀

1 reply

0 recast

1 reaction

0zAND1z pfp

life is based 🔵

0 reply

0 recast

1 reaction

0zAND1z pfp

oh hai. again. almost forgot to log that i have indeed started the ml pipeline implementation and have been updating the dataset/db services to appropriately support classification. @alchemyplatform to the rescue again! EOF. FR. 👾 3/n

0 reply

0 recast

0 reaction

0zAND1z pfp

also had fun shortlisting a bunch of libraries to render a visually appealing complete graph based on the combinations we have. intend to scratch the jupyter notebook tomorrow and share some samples here. for now, im cozy with matplotlib and d3.js. got ideas? pls drop your lib recommendations below. 👇 --EOF-- 2/n

1 reply

0 recast

0 reaction

0zAND1z pfp

DAY 6: a based and blessed morning to all! 🔵 today's buildlog from the flairsmith: i've compiled a list of labels, qualifiers, filters, criteria and color codes for historic external transactions. in total, there are 87 combinations just in this category and would love some feedback. DC me for access. 👀 1/n

1 reply

0 recast

1 reaction

0zAND1z pfp

0 reply

0 recast

0 reaction

0zAND1z pfp

i feel attacked with that AQI monitor 💀

0 reply

0 recast

0 reaction