Data Science and Machine Learning Podcasts

2023 May 25

Data Science and Machine Learning Podcasts

Posted on May 21, 2021

The subject came up in a thread on the Data Hackers forum and this interesting list emerged. Some of them I didn’t know personally, but here are the references for anyone interested.

Portuguese

Data Hackers: No further introduction needed as it is the official podcast of the largest Data Science, Machine Learning, and Data Engineering community in Brazil today; the podcast features a technical yet relaxed approach that goes beyond the commonplace and has many excellent episodes. It is hosted by universal Data Scientist Paulo Vasconcellos, Data Engineer of all solar system tools Allan Sene, and Gabriel “The DS Manager” Lages. I had the pleasant opportunity to participate in the episode “The day-to-day life of a Machine Learning Engineer — Data Hackers Podcast 24” and it was a sensational experience where I learned a lot from the guys and the recording was very fun (highlighting Lages’ totally random reference who cited the case of a lib that simply stopped being developed because the main developer was arrested).
Dados e Saúde: Whenever the subjects “Medicine” and “Artificial Intelligence” appear in some news, it’s always more of the same: A bunch of unfounded promises, hype, and a ton of nonsense to generate engagement. For those looking for a balanced view on the mixture of these two universes, the Dados e Saúde podcast is an island of sanity, pragmatism, and deliberation on these matters. Hosted by Fabiano Filho, Eduardo Farina, and Matheus Coradini, the podcast manages to bring realistic perspectives on how data is helping to improve people’s lives from a medical perspective. The podcast has an interesting set of episodes. Startups using data for innovation? They have them. Realistic information from those in the trenches applying AI in medicine? They have it. Challenges of applying Deep Learning to medical images? They have that too. Algorithms in medicine? No problem, they have them too. Highlight episode: Here I leave what I consider one of the best episodes, which was the chat with Professor Luis Correia from the excellent Evidence-Based Medicine blog in episode #47 Evidence-based rationality in healthcare — with Luis Correia.
Mario Filho: Mário Filho is one of those legendary figures in the Data Science community not only for his technical knowledge but also for the fact that he is a Kaggle Grandmaster (like, a real Kaggle Grandmaster who wins competitions and codes, not the current generation of KGM who fork code and beg for votes in forums). Along with Giba and the late Leustagos, they explored Kaggle with a 100% Brazilian team in mid-2013 (when Data Science didn’t even exist yet, imagine). Personally, I’m biased to speak about Mário’s technical capacity since something like 6 to 7 years ago he was applying *ensembling, stacking, *calculating similarity using *hashing *and all using XGBoost and Deep Learning; things that seem *mainstream *today but at the time went head-to-head even with academic publications in ML. Highlight episode: How I organized myself to complete more than 50 online courses
Pizza de Dados: I was a reader of Letícia Portella because of a very interesting post she made about traffic mortality in Brazil and by chance I ended up finding the Pizza de Dados podcast she does along with Jéssica Temporal and Gustavo Coelho. A cool thing I find about the podcast is that all the people from the technical community that I think “Wow, this person is awesome” have certainly passed through there. Highlight episode: Here the recommendation is episode 007: Ethics, laws, and data security which shows that the podcast is way ahead of its time regarding its agenda, given that this interview brought not only the ethical-legal perspective in 2018 long before the LGPD and the hype about AI ethics that exists today.
Databasecast: Hosted by Dr. Mauro Pichiliani and Wagner Crivelini, Databasecast is a grassroots podcast that talks about everything from backup *and *restore to data science and AI. One of the greatest highlights of the podcast, in my opinion, is the pragmatic way themes are treated; there is no enchantment with technology or hype, and both Mauro and Wagner speak the naked and raw reality of the daily life of professionals like DBAs: Backups fail, not everyone has Disaster Recovery, the data career is hard, among other things. The variety (and high caliber) of the professionals they invite are another highlight. Many of the stories I experienced firsthand (e.g., deleting a production database, update *without a *where, among other tragedies) I saw that even great professionals with vast experience had gone through the same things; and even with all that, they learned and shared their field experience with a very high degree of honesty. Episode DatabaseCast 85: Dismissal and unemployment was one of the best episodes I’ve ever heard, and it tells the naked and raw reality that usually doesn’t appear in Medium posts or company blogs when the axe falls.
Intervalo de Confiança: Founded by Nicolli Gautério and Patrícia Balthazar, with almost 2 years on the road, IC has an approach that I find interesting because they fill a very large gap on the Brazilian internet for true data journalism, something more or less along the lines of FiveThirtyEight. And it’s no wonder: the team has a very wide diversity of professionals and it’s very well produced and directed. If mainstream media networks had serious data reporting newsrooms, the result would be IC. Highlight episode: Variance — XIV — Artificial Intelligence and Crime Prevention
Data Cast: Made by Grupo Toccato, the podcast started in 2020 and its highlight is the light, accessible, and didactic language in approaching complex themes like digital transformation and LGPD. The guests in the episodes I heard have vast corporate experience, and with episodes of approximately 30 minutes, the podcast goes straight to the point. Highlight episode: #5 LGPD: What is still missing for your company to adapt to the law and how to accelerate this process.
Analytics Selvagem: The podcast from Oncase presents itself as “an initiative by Oncase with the objective of sharing, over several episodes of this Podcast, a bit of the daily life of professionals working with Analytics, Big Data, and Artificial Intelligence.” and that’s what the podcast delivers. Scripted and hosted by Lucas França and Henrique Tavares, the podcast talks about the daily life of a data consultancy and has some very interesting episodes like Episode 4: DataOPS, the buzzword.
Teste de Turing: Hosted by Erick Fonseca, in his own words, TdT is “Machine translation, chatbots, AIs that write like people, and so many other things: how are machines learning to speak our language? The Turing Test explains it all!”. Even though the title gives the impression of being a generalist podcast, TdT talks about implementation details and is extremely technical. A sample of the depth of the podcast can be found in the episode on syntactic *parsers *and Computational Linguistics. Highlight episode: Turing Test #5: Wordnets and Lexical Resources
Let’s Data Podcast: Hosted by Bernardo Lago, Felipe Schiavon and León Silva, the podcast started just now in March 2021, and seeing the excellent interview with Pedro Albuquerque you can already see that the podcast promises to be one of the best in the segment for those who like a well-scripted conversation, but without the commonplace of closed questions and answers. One thing I found interesting in the script is that Bernardo, Felipe, and León have a mixture of data knowledge from both the market perspective and a strong academic base.

English

So, what did you think? Is there any podcast you want to recommend? Leave it here in the comments because I want to get to know more Data podcasts.

…

Update 03/18/2021: Fix broken links.

Update 03/31/2021: Inclusion of Portuguese podcast descriptions.