Ape/rture

@aperture

109 Following

20 Followers

Ape/rture pfp

Slowly we are seeing a theoretical interpretation layer forming which interprets blockchain data, not only from a single chain but across multiple chains. The process is slow, since spinning up the full pipeline from indexer including processing can be tedious.

1 reply

0 recast

0 reaction

Ape/rture pfp

Data should be a first class citizen on the blockchain

1 reply

0 recast

3 reactions

Ape/rture pfp

Interested in devs and data engineers opinions on getting blockchain data: 1. How long does a backfill take (or your provider) until you can start adding realtime/tip of chain data 2. How expensive are these backfills? (price range should be enough) Helps to note which chains

1 reply

0 recast

0 reaction

Ape/rture pfp

In crypto we break through the concept of silos (data, IP, products). Traditional business shield most of these. Crypto enables new business models and products. A misstep crypto projects make is not experimenting with new models. They flee too quickly to a (badly designed) token, which hurts the project long term.

0 reply

0 recast

0 reaction

Ape/rture pfp

At some point we are going to see incentivization to have users connect more wallets to Farcaster. The proof of humanness Farcaster brings is through unique acts of humanness like: sharing thoughts, composing unique content and responding to and evaluating other humans.

0 reply

0 recast

0 reaction

Ape/rture pfp

When looking at data analyst or scientist job descriptions in crypto in a lot of cases I see that in the same function indexing, data engineering, ETL + analysis, insights and dashboarding is asked. In most cases these companies are searching for a unicorn and would be better to split engineer and analyst functions.

0 reply

0 recast

0 reaction

Ape/rture pfp

I love how easy streamlit is do to dashboarding. From idea to POC in just 2 hours

0 reply

0 recast

0 reaction

Ape/rture pfp

Amazing to see @meshconnectapi aligned with the @indexingco vision: "... Mesh can now access the critical data it needs without the substantial cost and overhead of building indexing infrastructure in-house." https://www.linkedin.com/pulse/unlocking-new-possibilities-crypto-transfers-mesh-partners-89kjc

0 reply

0 recast

1 reaction

Ape/rture pfp

One beautiful thing about being able to pull Farcaster data is that if you don't agree with the Farcaster Spam filter, you can build your own. By labelling the spam accounts by hand, there is a certain bias in the model too. User could prefer their own bias and could create spam filtering models themselves.

1 reply

0 recast

0 reaction

Ape/rture pfp

If you would want an API for Farcaster casts (mostly the text data), what data would you pull and for which types of applications?

0 reply

0 recast

0 reaction

Ape/rture pfp

I'm wondering how many data analysts in crypto prefer working with Python (for example Pandas) or Javascript (like D3.js). Most analysis is focused around SQL (if not on API calls), but personally feel this limits metrics I can build or analysis I can do like statistical analysis or even ML.

0 reply

0 recast

1 reaction

Ape/rture pfp

The cost of blockchain data is going to 0. Too many expensive data sets or services trying to provide a view, while the blockchain is a public database. This will make crypto actually usable, since we now can focus on higher level insights and spend dev time on the product.

0 reply

0 recast

1 reaction