Nuno Castro

Director AI Forward Deployed Engineering @Databricks

[Nuno Castro @nunodata]

Agentic development is already here, and it's got some demands for databases:
• Instant branching at zero cost
• True scale-to-zero (half of agent DBs live < 10 seconds)
• Open Postgres, not proprietary lock-in

#AgenticAI #Databricks #Lakebase https://t.co/KZ5IhgnCJf

[Ali Ghodsi @alighodsi]

@JeffDean says it best, the problem in this new agentic era is "tools designed for human speed interaction". That's why we think agents love 𝗟𝗮𝗸𝗲𝗯𝗮𝘀𝗲 𝗣𝗼𝘀𝘁𝗴𝗿𝗲𝘀, it can branch, snapshot, scale up and down in a second, orders of magnitude faster than other databases.
https://t.co/s32LICSEXc

Read about this architectural shift from Database to Lakebase:
https://t.co/hL19TVTIaH

[Nuno Castro @nunodata]

AI Papers to Read in 2020 by Ygor Rebouças Serpa in @TDataScience https://t.co/oOYvCbfPg1

[Nuno Castro @nunodata]

Pandas Version 1.0 is Out! Top 4 Features Every Data Scientist Should Know https://t.co/NkDecfj5u6 via @analyticsvidhya

[Nuno Castro @nunodata]

“14 Deep and Machine Learning Uses That Made 2019 a New AI Age.” by @Merzmensch
https://t.co/tobkjFQdU2

[Nuno Castro @nunodata]

@gigsandtours @coldplay @NHM_London You should have a queueing system in place not a random draw where basically time of arrival in the queue is irrelevant. The current one is a like a raffle, where you go back to the end of the queue on each refresh.

[Nuno Castro @nunodata]

@mac_isidoro @UKChange Yep this felt more like a draw than a queuing system

[Gerard de Melo @gdm3000]

Amazing!! Deep Learning-based NLP techniques are going to revolutionize the way we write software. Here's Deep TabNine, a GPT-2 model trained on around 2 million files from GitHub. Details at https://t.co/8J2v8Ns7n4 #nlproc

[Nuno Castro @nunodata]

Google launches new search engine to help scientists find the datasets they need https://t.co/EPPz6LmRhM via @Verge

[Christoph Molnar 🦋 christophmolnar.bsky.social @ChristophMolnar]

Explanation algorithms for neural network predictions are like yogurts in a supermarket. Choice overload.

LRP
LIME
Integrated Gradients
deconvnet
Occlusion
DeepShap
DeepLift
Deep Taylor Decomposition
Gradient* Input
Grad-CAM
Guided Backpropagation
Saliency Map
...

[Andreas Mueller @amuellerml]

Scikit-learn 0.21 is out! Including the new HistGradientBoostingClassifier/Regressor by @hug_nicolas that implements lightgbm and should match or improve over xgboost performance. Full changelog here: https://t.co/YZfvGg4XQt Also, there's now an option for a nicer __repr__!

[Nuno Castro @nunodata]

Google AI Blog:Introducing Translatotron: An End-to-End Speech-to-Speech Translation Model https://t.co/7dYS9rMAsk via @googleai

[Nuno Castro @nunodata]

As search needs evolve, Microsoft makes AI tools for better search available to researchers and developers https://t.co/wy0ee81j8y

[Microsoft Research @MSFTResearch]

“We've written this book to cover the theory we expect to be useful in the next 40 years, just as an understanding of automata theory, algorithms & related topics gave students an advantage in the last 40 years.”

Download “Foundations of Data Science” now https://t.co/dSlmNYajDw

[Elie Bursztein @elie]

Introducing Keras Tuner: hypertuning for humans - Slides: https://t.co/6vbyNtTti7 Early Access program: https://t.co/KUb3iTSiMk #io19 @TensorFlow @fchollet

[Nuno Castro @nunodata]

The ‘best job in America’ pays over $108,000 a year — and has a high number of job openings https://t.co/g8tIiBrCqW

[Nuno Castro @nunodata]

XGBoost Algorithm: Long May She Reign https://t.co/IbCelMxqfv

[Nuno Castro @nunodata]

Achieving a top 5% position in an ML competition with AutoML by @LarsHulstaert https://t.co/a0JoQPcCss

[eytan bakshy @eytan]

After many years in the making, we're open-sourcing Facebook's Adaptive Experimentation platform, Ax, and BoTorch, a library for Bayesian optimization research, built on PyTorch. https://t.co/1wpFnlqmPg

[PhocusWire @phocuswire]

How @expediapartners is using deep learning to improve the hotel booking process https://t.co/jY20YBBZaR via @Phocuswire / @foxychops