Edward Cui

Working on SaaS
Founder of Diffus.
Founder of Graviti.
Advocator for open data.
Ex-Software engineer and TL at Uber ATG.
Machine learning enthusiastic.

[Edward Cui @YunkaiCui]

Spending quite a lot on Google Ads I found out that computing the ROI of Google Ads for SaaS is quite different than a regular product!
For regular product is: Average Payment / Cost per Conversion
But for SaaS, it should be: Life Time Value / Cost per Conversion
#GoogleAds #SaaS

[Edward Cui @YunkaiCui]

This is what we build lately! It has been a fun journey so far. Get some extra credits towards your next master piece. #StableDiffusion #FreeCredits #Diffus https://t.co/XbQhqu1Lqq

[Edward Cui @YunkaiCui]

@dreamingtulpa @GravitiDiffus

[Edward Cui @YunkaiCui]

@ntraft Do we really understand what is the strength of this model at all at the designing phase? Or it is just a big model Open AI trained but letting us figure out what it can do well or not so well.

[Edward Cui @YunkaiCui]

What have I done that desires the purest smile from my daughter. I'm just being there, and she gave me the warmest and biggest small I have ever seen.

[Edward Cui @YunkaiCui]

I guess it will be super hard to beat 🦄 s. But they will drain money faster than my clicking anyway.

[Chris Bakke @ChrisJBakke]

If you’re a founder and you don’t click on 10 of your competitors’ ads every morning to slowly drain them of several dollars of ad spend and push them closer to bankruptcy $2 at a time, you’re a 1x founder not a 10x founder.

[Edward Cui @YunkaiCui]

At @graviti_ai , we are doing exactly that to accelerate both machine learning and data science workflow.

[Edward Cui @YunkaiCui]

Pandas DataFrame and Dask both use numpy as the backbone. It's Python native. However, when processing super large volume of data, columnar store has its strength. Why don't we reimplement DataFrame with a Apache Arrow backbone? And we can leverage the modern data infra. #Python

[Edward Cui @YunkaiCui]

The obvious advantage of using Arrow is that the data can be passed along to different services that are written by different programming languages. And we could possibly use different SQL engine to drive that query that runs on super big datasets. #ApacheArrow #DataFrame

[Edward Cui @YunkaiCui]

The storm is coming.

[Edward Cui @YunkaiCui]

Come to see my talk virtually or physically at Open Source Summit in Austin, TX. Thanks @linuxfoundation for providing such an amazing opportunity. #ossummit #openbytes #OpenData

[Graviti @graviti_ai]

Good morning everyone! Our founder @YunkaiCui will speak at Open Source Summit North America 2022 tomorrow, June 21st sharing the latest progress of Project OpenBytes under @linuxfoundation
Schedule the session with us:https://t.co/RIIbdghwqP #ossummit #OpenSourceSummit

[Edward Cui @YunkaiCui]

I like this answer very much. It extends to product development for software in general. Better designed software will have its APIs speak for itself.

[Musa Parmaksiz @mteucer]

@mierrashid @ylecun What makes a framework better is its better mental representation used to predict its behavior. IMHO TensorFlow had a huge disadvantage there, it was an UFO in Python world...

[Musa Parmaksiz @mteucer]

@mierrashid @ylecun What makes a framework better is its better mental representation used to predict its behavior. IMHO TensorFlow had a huge disadvantage there, it was an UFO in Python world...

[Edward Cui @YunkaiCui]

Coming back to Pittsburgh is quite fun! Finally be able to catch up with some old friends and see the development of the strip district.

[Edward Cui @YunkaiCui]

Just a fun fact. At @graviti_ai , 48.9% of all our FTE are females. 58.5% are females if interns are included. My talented coworkers have made us a better company! I have to cheer for it🌹 #InternationalWomensDay #womenintech

[Edward Cui @YunkaiCui]

@al_biglan And the cables/chargers in it.

[Edward Cui @YunkaiCui]

Pray for world peace🕊️

[Edward Cui @YunkaiCui]

We finally came out of stealth mode. A small step of a much longer journey.😎 https://t.co/975oYDrqDI

[Edward Cui @YunkaiCui]

Although my entire career is all about improving efficiency. Statistically speaking, efficiency sometimes will eliminate diversity. I am actually terrified to live in a world that is extremely efficient at the cost of sacrificing some of my random need.

[Edward Cui @YunkaiCui]

Sometimes, I am just wondering whether efficiency is good.

[Edward Cui @YunkaiCui]

For example, chain stores are optimized for efficiency, so they only have goods that most of the people want. Occasionally, when you want to buy something long tail, you will find yourself in the situation where you either have to order online or drive long distance.