The Artists of Data Science

The ONLY self-development podcast for Data Scientists

Chat Transcript from Data Science Happy Hours 14, Dec 18 2020

16:31:14 From Thomas Ives : Susan Walsh is in the ROOM!
16:31:25 From Susan Walsh : just!!
16:32:38 From Eric Sims : Unbalanced data and unbalanced data scientists?
16:33:26 From Akshay Mandke : Sometimes you need to take a few steps back to leap forward. I think it relates to balance
16:34:26 From Carlos Mercado : @ Eric whats up
16:34:37 From Eric Sims : Yo!
16:34:52 From George Firican : Hi everyone!
16:35:23 From Ashit Debdas : Hello everyone
16:35:32 From Greg Coquillo : Hi Team!!
16:35:33 From Thomas Ives : George!
16:35:34 From Ashen Rana : Hi all! Spilled a drink on my laptop right at the beginning :D
16:35:41 From Lalita : hello everyone
16:36:00 From Carlos Mercado : oh noooo @ Ashen upside down + dehumidifier + rice ??
16:36:04 From George Firican : Hope all is ok, Ashen
16:36:17 From Ashen Rana : Lol I might have to dry the rice technique Carlos
16:36:24 From Carlos Mercado : @ Greg whats up man, got some more stump us questions this week?
16:36:32 From Carlos Mercado : @ Mark whats good bro
16:36:34 From Ashen Rana : Managing it for now, George - thanks! Seems okay for now thankfully
16:36:54 From Naresh Reddy : Hello everybody!
16:37:04 From Eric Sims : Hey @Naresh!
16:37:04 From Greg Coquillo : lol @Carlos I'm more on the learning flow today. Might have some questions
16:37:32 From Carlos Mercado : I asked a PhD in biophysical chem that question of yours on AlphaFold, still waiting for the reply, will send that answer to you on LI.
16:37:40 From Mark Freeman : @Carlos hey! Happy to see you in office hours this week!
16:38:05 From Harpreet Sahota : If anyone has questions - send me a message!
16:38:07 From Carlos Mercado : Trying do every other week at a minimum. These are so big now!
16:38:07 From Greg Coquillo : @Carlos that would be awesome! thank you
16:38:11 From Dave Langer : I would offer that analytics is like any other applied STEM field in business (e.g., software engineering).
16:38:41 From Dave Langer : That is, to stay relevant you need to constantly invest in your skills - and that usually happens outside of work hours.
16:38:51 From Saurabh Dixit To Harpreet Sahota(privately) : For someone starting new…. and looking at so many branded Data Science programs … :
16:39:28 From Dave Langer : For many, this level of consistent effort/investment year after year is too much sacrifice. If you love it, however, it's awesome.
16:39:30 From Ashen Rana To Harpreet Sahota(privately) : What are some trends that you’ve noticed on LinkedIn recently? Did you take a course on “how to LinkedIn”?
16:39:42 From Joe Reis : ^ this
16:39:57 From Saurabh Dixit To Harpreet Sahota(privately) : Tableaue Data Scientist, SAS Data Scientist, Microsoft Azure Data Scientist programs… .. how relevant and useful they are in the industry ..is it worth pursuing any of these
16:40:52 From Eric Sims : Welcome, Saurabh!
16:40:55 From Dave Langer : Just want to do a shoutout - Ameya in the house!
16:41:22 From Ameya Dhaygude : Thank you Dave :)
16:41:29 From Ameya Dhaygude : Hi Everyone
16:41:52 From Carlos Mercado : I would just remind people reading this later; that it doesn't HAVE to be outside work hours. People tend to grind their jobs, never saying no, and then get surprised that they have no time.

If your job benefits from you being up to date on skills and knowledge, then its PART of your job to do that. Whenever possible ,carve out a few hours a week of self-time, locked down time on your calendar, to read newsletter, read papers, etc. And if it benefits your job then it counts as your job.
16:42:04 From Mark Freeman : I have question on people’s process for debugging their code efficiently? What’s your personal order of operations as a flexible guide, or any tools you use?
16:42:06 From Florin Badita - Corruption Kills : One of my pet hobby project is scrapping a list with most of the users on the Interert. I currently have around 500M users
16:42:11 From Carlos Mercado : I agree with you @ Dave, just adding my consulting feeling.
16:42:15 From Akshay Mandke : Role based certifications by Microsoft are really great.I recently did one and I can concur it is beneficial to apply your skills.
16:42:47 From George Firican : That's impressive, Florin
16:42:55 From Florin Badita - Corruption Kills : My question was if you know scientists working on this, my aim would be to be able to do hobby based discovery of persons
16:42:56 From Ashen Rana : Good point, Carlos!
16:43:54 From Jennifer Nardin To Harpreet Sahota(privately) : Might start a war but…. As a business-side-data-geek, I want to deep-dive into either R or Python over vacation. Which should I choose? (Organization doesn’t care, nearly any tool available) I would ask for responders to give me 1 reason why to choose their language of choice and 1 reason why it should not be the other language.
16:43:54 From Akshay Mandke : Tools are simply enablers for data science.I think it begins with what problem you want to solve and do you have the 4Vs of data answered well before starting your DS journey.
16:44:16 From Ameya Dhaygude : Question for the experts - I am a SAS veteran, using it for last 10 years. My company stopped using SAS and moving to Azure. Can you share your experience and advice on managing this transition?
16:44:37 From Jennifer Nardin To Harpreet Sahota(privately) : If you’ve covered that, or if it creates contention, then I am happy to just choose offline :)
16:44:46 From Carlos Mercado : @ Mark on debugging. There are 3 levels of debugging (and probably more if you are legit in software development).

1 - Commenting out code and trying again
2 - formal debugging with traceback with something like browser() in R to change your environment
3 - formal unit testing (including automated testing) that return set outputs, e.g. unexpected classes, failed asserts, etc.

Maybe there's a 4th but that's my typical range.
16:45:04 From Matt Housley : My perspective is that R is a really nice transitional language from SAS into the data science mainstream.
16:45:32 From Srivatsan Srinivasan : Hi all.. Good to see you all in one place :)
16:45:40 From Eric Sims : Someone needs to make the "Data Science Unicorn Unicourse" - One course to rule them all...
16:45:44 From Akshay Mandke : Yeah I like R’s functional approach. You can layer around your code as you explore the data. Great if you are a statistician or exploring data for sales and making charts using ggplot.
16:45:59 From Greg Coquillo : +1 @Srivatsan!
16:46:10 From Carlos Mercado : @ Ameya

SAS -> Azure is a bit of a strange way of wording it. SAS the statistical language is easy to swap into something like R. SAS the integrated development workflows with SAS Vidya, etc. switching to Azure is like switching to a full cloud infrastructure. You have to care about VMs, database architectures, Containers (Kubernetes), etc. I think we'd need more info.
16:46:23 From Akshay Mandke : Hi Greg, good to see you.
16:46:56 From Carlos Mercado : SAS code just works, always, it's all backtested. In an open source programming language, dependency management starts mattering in ways that SAS just "works" for you.
16:47:00 From Greg Coquillo : You too AkshaY!
16:47:04 From Matt Housley : Once you know R, various data frame paradigms across different frameworks will make sense. Spark, Python/Pandas, etc.
16:47:38 From Mark Freeman : @Carlos unit tests have been REALLY helpful for my current project. I’m currently struggling with reading the error statements and stack traces when it’s not obvious.
16:47:48 From Eric Sims : Don't post links in your own posts!!!
16:47:50 From Carlos Mercado : that's because you're not using FP lol
16:48:14 From Carlos Mercado : @ Eric. Correct, anything that makes your audience leave LinkedIn is suppressed. The algorithm is always changing.
16:49:33 From Ashen Rana To Harpreet Sahota(privately) : What are your expectations from someone in a Junior role?
16:49:52 From Giovanna Galleno : your style is amazing Susan! I love your tone of voice! You rock! :D
16:49:58 From Carlos Mercado : yeah, to Dave's point. The fundamentals transfer - SQL isn't going anywhere. Designing databases isn't going anywhere. OOP/FP isn't going anywhere. Understanding debugging, classes, O Notation, etc.

16:50:19 From Akshay Adlakha : Agreed Carlos
16:51:44 From Carlos Mercado To Harpreet Sahota(privately) : You can remind people who are just chilling, that the chat is very active and answers will be recorded in the show notes. Lots of questions in here.
16:51:59 From Saurabh Dixit : Great point ..Susan, Giovanna … Be yourself .. fail, learn .. Thanks.
16:52:09 From Matt Housley : I would argue the SQL has grown more important. In the early days of Hadoop, you had to write your own raw map reduce jobs in Java. The advent of Hive changed that, and now SQL is an extremely powerful language at scale. SQL is still annoying sometimes, but it’s a baseline skill.
16:52:16 From Lalita : I have one question regarding landing a entry level machine learning or data science position.
16:52:26 From Eric Sims : Remember, Harpreet interviewed Greg before he was famous!
16:52:26 From Dave Langer : Regarding LinkedIn I would suggest the first thing to do is to determine why you are doing it. Growing a LinkedIn audience is a lot of work. For example: Do you want to establish your "authority" in a particular domain? If so, guide your content accordingly.
16:52:27 From Ray Givler : The other facet is data viz. Human perception isn't going to change any time soon. Clarity won't go out of style.
16:52:41 From Florin Badita - Corruption Kills : Carlos, totally agree . Just started reading Rapid Development by Steve McConnell, written in 1994, and the book is so up to the point and still relevant 25 years later
16:52:43 From Saurabh Dixit : Thanks Srivatsan and Dave… point taken.. focus on foundational skills, transition on the basis of domain knowledge… Thanks a lot.
16:52:44 From Giovanna Galleno : Thank you Saurabh :D
16:53:24 From Ray Givler : @Florin. I read that book in 1994!
16:53:25 From Carlos Mercado : My #2 post ever on LinkedIn was "Data Viz is more important than Deep Learning".
16:53:32 From Carlos Mercado : I'm a big fan of DV.
16:54:32 From Naresh Reddy : what transition does it take to become a data scientist from data analyst?
16:54:36 From Matt Housley : @Carlos Yeah, I second that. We’ve seen that issue with clients - hyper competent data scientists who struggle to communicate with execs because they can’t do data viz.
16:54:47 From Monica Royal : Don't fall into the LinkedIn algorithm trap. Just be you and share things that you enjoy. Also, remember we are all learning from each other, so anything that you post will help someone else :)
16:54:51 From Eric Sims : It's true - Greg never misses!
16:55:00 From Carlos Mercado : Can't do Data Viz and CANT DO POWERPOINT. Execs only speak 3 languages. English, PPT, Finance Excel.
16:55:10 From Susan Walsh : great point Monica!
16:55:12 From Greg Coquillo : Thanks Eric!
16:55:13 From Carlos Mercado : @ Naresh - you just call yourself a data scientist and you're done.
16:55:28 From Carlos Mercado : The roles are supposed to be distinct and different, but they're just not anymore, it's all fumbled.
16:55:38 From Eric Sims To Harpreet Sahota(privately) : Harpreet, I've got a technical question if I can get in the queue
16:55:46 From Mikiko Bazeley : OH man, debugging, good one
16:55:49 From Naresh Reddy : @carlos :D
16:55:52 From Matt Housley : @naresh One of the main things you need to make this transition is strong sponsor. Think of it as an internship.
16:56:34 From Naresh Reddy : @Matt Housley I will keep that in my mind :
)
16:57:18 From Carlos Mercado : At RStudio 2020 I took Jenny Bryan's workshop on debugging; she also gave a keynote speech on debugging. Available here:

https://rstudio.com/resources/rstudioconf-2020/object-of-type-closure-is-not-subsettable/
16:57:19 From Florin Badita - Corruption Kills : Thanks for organising this, I will need to sop, have another call coming up. Happy and relaxing xmas to you and your family. Stay safe!
16:57:31 From Eric Sims : I start worrying when I'm not getting enough errors...
16:57:47 From Ashen Rana : Echoing the importance of SQL. I thought it was too simple and easy, and not as exciting as Python or R. I am currently using DBT (data build transform) tool and SQL is key there
16:58:10 From Giovanna Galleno : byw @Florin! :D it was great to see you! :D
16:58:21 From Joe Reis : Check out my next meetup about ML model testing and debugging
16:58:22 From Joe Reis : https://www.meetup.com/utah-data-engineering-meetup/events/275230885/
16:58:27 From Eric Sims : See you, @Florin!
16:58:33 From Joe Reis : Josh Tobin (Open AI) is presenting
16:58:44 From Matt Housley : @ashen +1 for DBT.
16:58:47 From Akshay Adlakha : We can checkpoint while building a model. This makes easy to debug a model.
16:58:53 From Carlos Mercado To Harpreet Sahota(privately) : I have something I want to chime in on RE Debugging
16:58:58 From Akshay Adlakha : add*
16:59:05 From Akshay Mandke : Keeping in mind data privacy and restrictions for compliance plays a huge role in making changes to data pipelines and solutions. I have experience working in Forensics industry and see this as a big challenge.
17:00:11 From Carlos Mercado : In R the Pins library can record datasets, models, and objects that Srivatsan is mentioning. In Python I think Metaflow does caching for it too
17:00:56 From George Firican To Harpreet Sahota(privately) : I can mention a quick thing about debugging if there's time
17:00:57 From Ashen Rana : Heck yeah, Matt Housley! Getting started with DBT + Snowflake
17:01:38 From Ray Givler : For actual error messages, I like to record the message and my resolution and keep that in a document or OneNote. Multiple problems can cause the same error, so once you find out what yours is, it's worth writing down. If you make an mistake once, you are likely to repeat it. So keep some notes.
17:01:40 From Lalita : I think commenting the sections have helped me a lot while debugging my scripts in Python
17:01:44 From Joe Reis : Also, debug your data...
17:01:47 From Matt Housley : This framework is pretty interesting for model monitoring. https://github.com/whylabs/whylogs-python
17:02:15 From Susan Walsh : nice hat Monica!
17:02:26 From Joe Reis : Why Labs is great for data profiling in your pipelines. Great expectations is a good data unit testing framework
17:02:52 From Joe Reis : Data’s different than software debugging, in that you need to debug your software code AND your data
17:03:32 From Matt Housley : One of things you have to think about in data science debugging is looking for errors in your source data.
17:03:56 From Monica Royal : Thank you @Susan :D
17:03:59 From Carlos Mercado : Also RE: debugging. Take a break and come back if it's really bad. Then check your assumptions.

The most common debugs are the ones we don't post to stackoverflow. "Shit my date is being stored as a character".

It's like very basic CompSci stuff.

But the #1 way to debug is to create REPREX and actually isolate your error. That will solve it 80%+
17:04:35 From Susan Walsh : step away for a while then come back also helps
17:04:36 From Matt Housley : It really helps to track down domain experts in your company. They can often tell you when events or statistics don’t smell right.
17:04:38 From Thom Ives : Atom
17:04:38 From Dave Langer : RStudio!
17:04:39 From Ashit Debdas : pycharm
17:04:39 From Carlos Mercado : RStudio is the best Python IDE - don't @ me.
17:04:41 From Mark Freeman : Vscode with mypy and flake8!!!
17:04:42 From Greg Coquillo To Harpreet Sahota(privately) : I have a question as well brother
17:04:43 From Joe Reis : VSCode and vim
17:04:48 From Liuna Issagholian : VScode
17:04:49 From Kamarin Lee : VSCode +1
17:04:50 From Lalita : spyder
17:04:57 From DavidTello : I once missed a Space in a Python code, it took me all night to find it
17:04:59 From Monica Royal : Good point @Joe. You may just need to take a walk and come back to it later. Often times you will find the missing comma right away

17:05:02 From Saurabh Dixit : Sometimes just stepping away from your computer for 10 mins and then returning helps solve the problem ;) in my humble exp
17:05:05 From Ray Givler : @Carlos - I was gonna say the same thing regarding debugging. Assumptions are a common source of bugs.
17:05:22 From Akshay Adlakha : Guys, due to some urgency, I have to go. Bye Happy holidays. Enjoy rest of the session.
17:05:27 From George Firican : sorry everyone, I have to step out as I have a couple more hours in the office before the weekend starts. Lovely seeing everyone here
17:05:28 From Joe Reis : If I had a dollar for every simple bug caused by a misplaced character, I’d be retired
17:05:30 From Susan Walsh : Florin that is a wicked camera you have
17:05:34 From Eric Sims : Take care, @Akshay!
17:05:34 From Joe Reis : Later George
17:05:35 From Matt Housley : Has anyone debugged LaTex code? It’s horrible.
17:05:44 From Joe Reis : Matt - that’s morbid
17:05:50 From Akshay Adlakha : Thanks Eric
17:06:00 From Susan Walsh : Bye George!
17:06:09 From Liuna Issagholian : LaTex …… can be challenging :|
17:06:14 From Mark Freeman : Thank you everyone for your answers! This was so helpful and will coming back to this podcast for all the great gems on debugging
17:06:26 From DavidTello : @Matt, I wrote my dissertation (120 pages) using latex. It was painful
17:06:31 From Mark Freeman : https://osintframework.com/
17:06:43 From Mark Freeman : ^ open source Carlos was referring to
17:07:02 From Eric Sims : Sounds YUGE.
17:07:39 From Thom Ives : Mikiko's antlers are EPIC!
17:07:51 From Carlos Mercado : +1 Mikiko's antlers
17:08:19 From DavidTello : Is web scrapping illegal?
17:08:29 From Joe Reis : no
17:08:29 From Carlos Mercado : It's not GDPR compliant
17:08:33 From Carlos Mercado : in most instances
17:08:43 From Carlos Mercado : because you would be liable for certain types of identifiable storage.
17:08:45 From Lalita : Is there any good resource where I can learn writing complex queries in SQL problem just to play around with it more?
17:08:51 From Thom Ives : I'm going to be arrested if it is!
17:08:55 From Joe Reis : If you’re gonna scrape, just keep it quiet
17:09:05 From Dave Langer : ^
17:09:10 From Sarah Nabelsi : lol
17:09:10 From Thom Ives : Lalita - David Langer's course!
17:09:13 From Susan Walsh : we’ll come visit Thom
17:09:21 From Matt Housley : @DavidTello I think it lands in a legal gray area. You’re typically violating terms of service, which you may or may not have actually signed off on.
17:09:44 From Thom Ives : I dream of it Susan!
17:09:49 From Dave Langer : Lalita - I have a free SQL tutorial if you're interested.
17:09:52 From Lalita : Thanks Thom
17:10:05 From Eric Sims : "I don't intend to start a war" - Famous last words immediately before starting a war.
17:10:13 From Susan Walsh : lol
17:10:20 From Lalita : @Dave thanks. I would love to use that.
17:10:28 From Mark Freeman : I’m a heavy python user… but I think R is more friendly to have it just work
17:10:31 From Mikiko Bazeley : For scraping, there is some many conditions
17:10:34 From Mikiko Bazeley : C++!
17:10:41 From DavidTello : Flip a coin and let mathematics decide :)
17:10:49 From Akshay Mandke : R if only researching, python if its part of a larger software solutions
17:11:01 From Susan Walsh : I don’t code 😱
17:11:04 From Matt Housley : @Lalita One nice resource is the DBT walk through. They really emphasize using CTEs to compose complex operations. Basic composition is one of the most under appreciated capabilities of SQL, largely because business users don’t use it, and we tend to learn from business users.
17:11:13 From Dave Langer : @Lalita - https://www.daveondata.com/blog/sql-for-excel-users-part-1/
17:11:46 From Akshay Mandke : R is easier if u have no programming backgroun
17:11:53 From Joe Reis : I’m learning swift right now…
17:11:55 From Susan Walsh : see which one you get on best with
17:11:59 From Joe Reis : It’s amazing
17:12:11 From Kamarin Lee : plotly ftw lol
17:12:16 From Dave Langer : In my experience teaching 100s of working professionals, R is easier to learn for folks with no programming background.
17:12:20 From Lalita : Thanks @Matt and @Dave I will check that out
17:12:24 From Sarah Nabelsi : +1 to harpreet
17:12:26 From Carlos Mercado : The V8 package uses Chromium to render JS on webpages prior to scraping (this is unlike rvest in R that has issues with rendering JS first).

You will find a LOT of sites actively prevent webscraping; so in terms of being a white hat hacker. You should protect your sites from scraping by making sure to render important details as JS outputs and not HTML, and also make sure it can effectively block V8 / Chromium engine rendering of JS.
17:12:27 From Eric Sims : @Jennifer - I'm a noob, and I started with Python. It works well for me, and it is widely used in companies that I want to work with, so it's a skill I want.
17:12:29 From Faraaz Sheriff : I believe the learning curve for R is steep vs Linear for Python. Agree?
17:12:41 From Susan Walsh : Hey @Kam
17:12:49 From Joe Reis : Let’s put it this way - my 10 year old can write Python
17:12:52 From Saurabh Dixit : I was a Java developer long back .. and I am now starting to love R ..it’s amazing how much data wrangling it does so easily… .. No strong feelings .
17:12:55 From Dave Langer : If you're working a a solo analyst it is hard to go wrong with R. If you need Production software engineering, then Python might be a better choice.
17:13:06 From Carlos Mercado : I think R step 0 is hilariously easy. And Python step 0 is horrendous. But Python is very readable and has I believe 10x+ the stackoverflow answers.
17:13:10 From Kamarin Lee : @Susan!!! It’s been too long, great to see you, lovely! I’m in love with the Xmas edition you produced today haha :)
17:13:23 From Carlos Mercado : My opinion: R for individuals; Python for Teams.
17:13:28 From Eric Sims : @Carlos - R doesn't have a step 0. Python starts at 0.
17:13:33 From Carlos Mercado : ha ha ha LOL
17:13:35 From Susan Walsh : Haha thanks, that was freshly produced today!
17:13:46 From Karan Ambasht : Sorry gotta run.. great listening to wonderful thoughts from everyone. Happy Holidays all !!
17:13:54 From DavidTello : I started in Python, moved to VBA because of work needs
17:13:55 From Carlos Mercado To Harpreet Sahota(privately) : Can I shout out my mentee on this conversation really quick
17:13:59 From Lalita : I find Python my to go language. Never touched R
17:14:12 From Matt Housley : Python is a true object oriented language, so there’s a steep learning curve. R doesn’t have that problem, but won’t help you to segue into Java like Python will.
17:14:14 From Eric Sims : To Carlos' point, I love Python because it has a b'zillion StackOverflow answers
17:14:18 From Kamarin Lee : @Susan I need to think of ideas for a duet next year haha… 99 Luftballoons
17:14:19 From Joe Reis : Python’s the 2nd best language at everything
17:14:29 From Carlos Mercado : ^ +1 It's truly general purpose.
17:14:47 From Dave Langer : Ben!!!!
17:15:00 From Susan Walsh : Hey Ben!!
17:15:00 From DavidTello : If you want to do R, check out Matt’s courses, among the best I seem
17:15:16 From Greg Coquillo To Harpreet Sahota(privately) : Le me know if I'm in line for a question
17:15:16 From Carlos Mercado : I personally think R has the better community as well, which does matter for learning.
17:15:35 From Matt Housley : +1 for Matt Harrison’s Python courses.
17:15:40 From Ameya Dhaygude To Harpreet Sahota(privately) : I have a question on SAS to Azure transition
17:15:48 From Mikiko Bazeley : +1 for Matt Harrison
17:15:49 From Saurabh Dixit : I had the same Q
17:15:52 From Mikiko Bazeley : And his course
17:16:10 From Joe Reis : Matt Harrison is the king of Python courses
17:16:11 From Eric Sims : Hey, @Al! I hope things are less crazy this week than last!
17:16:23 From Kamarin Lee : Python has been instrumental for me with automation + analytics - Matt Dancho is amazing with R though. Definitely recommend his courses
17:16:59 From Akshay Mandke : +1 ^ I am doing his DS4B-101
17:17:04 From Akshay Mandke : in R
17:17:08 From Carlos Mercado : Python also has the excellent book "Automate the Boring Stuff". For command line automations Python is just unmatched.
17:17:32 From Joe Reis : I think of R like a hand calculator. It’s awesome for out of the box stats
17:17:33 From Ashen Rana : +1 to documentation. That’s been my emphasis
17:17:45 From Matt Housley : Right on Srivatsan
17:17:45 From Ashen Rana : I see you shaking your head, Dave lol
17:17:56 From Dave Langer : SQL, baby!
17:17:56 From Susan Walsh : lol
17:17:57 From Kamarin Lee : Data Storytelling - documentation and translating insights from reports is also incredibly valuable. I personally look for SQL + data storytelling ability before Python
17:18:01 From Mark Freeman : I LOVE SQL SO MUCH!!!!!
17:18:05 From Sarah Nabelsi : Dave!
17:18:06 From Sarah Nabelsi : haha
17:18:08 From Sarah Nabelsi : All you
17:18:10 From Ameya Dhaygude : Super agree with Srivatsan on SQL
17:18:27 From Akshay Mandke : CTEs are life savers
17:18:32 From Joe Reis : Also underrated - shell scripting
17:18:42 From Susan Walsh : next question - how do you say SQL?? 😂
17:18:42 From Carlos Mercado : in R you can do.
summary(lm(y ~ x)) and immediately start talking about model outputs. completely out of the box.

Also R has tidyverse; an entire optimized dialect that even has a SQL converter with dbplyr.

Like RStudio makes R worth learning and the community is insane.
17:18:45 From Ameya Dhaygude : Thank you Srivatsan for mentioning the jargon
17:18:46 From Sarah Nabelsi : I wanna also say creativity!
17:18:47 From Kamarin Lee : the industry conflating ML with AI for example…LOL
17:18:53 From Sarah Nabelsi : YES!
17:19:05 From DavidTello : Does anyone know of a SQL course that covers questions of the type that Eric Weber usually shares
17:19:21 From Ashen Rana : 3 months into my Jr. Data Developer role and I believe I got this role mostly because of my soft skills hah
17:19:25 From Dave Langer : For the teams that I've managed, Jr. Analytics folks are SQL, R (Python is OK, I would teach you R), and some basic knowledge of data analysis.
17:19:33 From Eric Sims : @DavidTello, Eric Weber also shares lists of helpful SQL courses, so definitely check them out!
17:19:59 From Matt Housley : @akshay I think that CTEs have developed a bad reputation because people don’t document them. Basically, SQL is a great language in its domain if we treat it like a programming language and actually document and version control.
17:20:00 From Carlos Mercado : Things a Jr Data Scientist needs to have:

  • Be curious
  • find interesting relationships and BRING THEM TO PEOPLE - don't over-dive in. Get team input
  • Good data viz
  • Can talk to people, i.e. with Data Viz and PPT
  • Python or R
  • Can handle their own basic debugging and stackoverflowing to a reasonable level. 17:20:08 From Harpreet Sahota To Greg Coquillo(privately) : Hey! Yea I have you after two more- feel free to jump in on any response though 17:20:26 From Carlos Mercado : Better to understand GLM very well; then know 10 different ML models and have no clue how to use them. 17:20:42 From Akshay Mandke : @matt I totally agree. I have struggled fixing broken scripts and CTEs that don’t work in my journey. Documentation is crucial. 17:20:42 From Ashen Rana : Good points Carlos! It’s surprising how many people/teams work in silos and bringing those people together is a skill by itseld 17:20:48 From DavidTello : Thanks @Eric Sims 17:20:51 From Kamarin Lee : Jr. Data Scientists should know how to translate data analyses with sound data storytelling and collaborative abilities 17:22:03 From Carlos Mercado : Hey @ Everyone, my mentee William Rodriguez just graduated from college; I mentored him in R, he's looking for his first analyst role either remotely or in the Orlando area. It'd be awesome if y'all added him on LinkedIn and could help him out. 17:22:09 From Carlos Mercado : He's in the chat today 17:22:34 From William Rodriguez : Hey everyone! 17:22:37 From Matt Housley : @akshay My standard is that someone should be able to look at a subquery in my CTE and understand what it does based on the inline comments. We generally expect that for functions in Python, but don’t always apply that standard in SQL. 17:22:44 From Eric Sims : Hey @William! 17:23:02 From Giovanna Galleno : Hola William! :D 17:23:32 From Ashit Debdas : Hello @william 17:23:33 From Akshay Mandke : @matt concur. and don’t forget nested sub queries. That’s where CTEs really play a role especially in complex analytics for financial datasets. 17:23:55 From Dave Langer : Death to nested sub-queries! 17:24:04 From Joe Reis : ^ amen 17:24:06 From Matt Housley : Yeah, nested subqueries can be horrifying. 17:24:28 From Akshay Mandke : Are they not so popular from your experience? I want to know if I need to avoid them in future 17:24:45 From Greg Coquillo : Tell us how you like nested sub-queries @David 17:24:48 From Greg Coquillo : lol 17:24:51 From Joe Reis : They’re popular for all the wrong reasons 17:24:56 From Dave Langer : Also some heresy here. Make all your SQL code CAPS for the love that all is holy! :-p 17:25:11 From Susan Walsh : 😂😂 17:25:11 From Dave Langer : The reserved words, that is. 17:25:16 From Ashen Rana : Hi Al(bert) Bellamy! Nice to see you here :D 17:25:20 From Carlos Mercado : I don't SQL and nested subquery just sounds bad as a word lol. I would assume never do it. 17:25:26 From Matt Housley : Depends - I’ve just seen some horrible examples of nested subqueries that would be much nicer in a CTE 17:25:45 From Matt Housley : But I generally don’t have a problem with them if they’re readable. 17:25:46 From Carlos Mercado : All my SQL skills are: Select * FROM x WHERE … Then save to csv and work in R LOL. 17:25:49 From Srivatsan Srinivasan : CTE has both sides.. But if it is analysis sometimes we might have to just iterate over same base query and generate different rollups quickly 17:26:13 From Akshay Mandke : nested sub Qs and CTEs if not designed well can throw your script into a never ending loop. 17:26:28 From Dave Langer : @Carlos - LOL 17:26:40 From Akshay Mandke : @shrivatsan - great advice 17:26:50 From Mikiko Bazeley : The limit for nested subQ’s & CTE’s was wholly dependent on how loudly our CTO yelled in the data warehouse slack channel 17:27:10 From Joe Reis : It’s the same philosophy in Python - most devs spend their time reading code, not writing it. Make your code readable. 17:27:13 From Greg Coquillo : lol @Mikiko 17:27:17 From Susan Walsh : I need to bail guys, need to sleeeeeep. Great to see you all! 17:27:25 From Joe Reis : Later Susan 17:27:27 From Eric Sims : Sleep tight, @Susan! 17:27:34 From Dave Langer : Bye, Susan! 17:27:38 From Susan Walsh : cheerio.... 17:27:39 From Greg Coquillo : Bye Susan! 17:27:41 From Ashen Rana : Nite Susan ! 17:27:42 From Giovanna Galleno : bye Susan! It was AMAZING to see you here! :D 17:27:43 From Monica Royal : See you later Susan! 17:27:44 From Thom Ives : Bye Susan! 17:27:53 From Ashit Debdas : bye susan, take acre 17:27:54 From Kamarin Lee : See you soon Susan! Happy Holidays! 17:27:56 From Ashit Debdas : care 17:28:39 From Mark Freeman : I have to get back to work. But I enjoy this so much! Happy holidays! 17:28:50 From Eric Sims : Later, Mark! 17:28:55 From Greg Coquillo : same to you Mark! 17:29:02 From Joe Reis : See ya Mark 17:29:16 From Monica Royal : See you later Mark! 17:29:44 From Carlos Mercado : @ Lalita -> What is your sales pitch? -> Are you doing a deep application or doing spray and pray? -> I recommend the following: https://www.linkedin.com/feed/update/urn:li:activity:6695094810270863360/

I recommend deep applications; when you apply, make yourself stand out, break the rules.
17:29:56 From Carlos Mercado : Lalita, you can send me your resume, happy to review.
17:30:33 From Saurabh Dixit : It would be worth getting your CV reviewed and formatted by profession resume developers.. I say this because there are some filters at the entry stage which could potentially filter out the CV if some key words are not seen.
17:30:43 From Ashen Rana : Ben Taylor, what’s the story behind #unrecruitable on your LinkedIn? [thinking face emoji]
17:31:05 From Ben Taylor : I can talk finally! :)
17:31:17 From Thom Ives : Yay Ben!
17:31:23 From Ben Taylor : #unrecuitable... yeah there’s a story there
17:31:25 From Matt Housley : I would recommend specifically building relationships with data scientists and engineers who have an academic background. That’s how I got my first DS job, and that’s true for many of my friends with PhDs in math as well.
17:31:26 From Lalita : Thanks @Carlos and @Saurabh. I will do that
17:31:31 From Dave Langer : Yes! Ben in the house!
17:31:33 From Eric Sims : @Lalita - Being active on LinkedIn has helped me create authentic relationships that I really value. It has also helped me connect with potential job opportunities.
17:32:08 From DavidTello : 100% agree @Matt Housley
17:32:52 From Kamarin Lee : its about how YOU can contribute to the BOTTOM LINE for a business
17:33:02 From Sarah Nabelsi : YES! There we go
17:33:16 From Sarah Nabelsi : Do your research, find out how you can add value
17:33:18 From Sarah Nabelsi : And lead with that
17:33:39 From Carlos Mercado : Have a day 1 contribution plan. "I understand your business, I did research on you, I know you make money by X, I can add to that with [hire me to find out]".
17:33:41 From Thom Ives : CARE - sorry - 4 letter word!
17:33:43 From Ameya Dhaygude : @Lalita - It's also difficult to get an interview call if you will need visa sponsorship. You may be a top candidate based on your profile, but the visa requirements sometimes plays against international students. I have heard from several international students that active and persistent networking has helped them more to secure jobs. That's how I was able to get my job.
17:33:44 From Kamarin Lee : 👏🏽👏🏽👏🏽 @Mikiko amazing!
17:33:44 From Saurabh Dixit : Personally I recommend topresume.com .. I leveraged their service to reformat my CV … At a stage when I was laid off and was looking for new job. .. my personal experience there. .. we can’t do much if our CV is kicked out by automated systems
17:34:03 From Joe Reis : Just call them and yell, “DO YOU KNOW WHO I AM?!!!”
17:34:03 From Greg Coquillo : +1 Mikiko!!
17:34:07 From Joe Reis : Actually, don’t.
17:34:28 From Ashen Rana To Harpreet Sahota(privately) : Can we hear Ben Taylor’s story behind the #unrecruitable hashtag? :D
17:35:23 From Eric Sims : Greatest snow on earth.
17:35:24 From Mikiko Bazeley : OMG that’s amazing\
17:35:25 From Carlos Mercado : Unethical Interview tip:
Call the corporate number; and say the data science team didn't call you at 3PM last Thursday when they said they would and why. Benefits if you can name-drop a data science manager.
17:35:53 From Joe Reis : Brighton is a great resort
17:36:13 From Saurabh Dixit : Got to jump off.. .. Awesome audience , great advise.. thanks so much everyone.. !! Thank you Harpreet for connecting with so many amazing people in the field .!!!
17:36:22 From Eric Sims : See you, @Saurabh!
17:36:53 From Carlos Mercado : Just for the comment recording. Ben is actually taking this call from the ski slopes in Utah and answering this in full ski gear XD
17:36:58 From Saurabh Dixit : Cheers @Eric, all
17:36:58 From Mikiko Bazeley : OMG Ben were you snowboarding while on the call!!! #FridayGoals
17:37:29 From DavidTello : while sky diving :)
17:37:32 From Carlos Mercado : 33 to 6 :( :( :( haha
17:37:38 From Joe Reis : While rodeo flipping
17:37:40 From Mikiko Bazeley : Besides Dave and Carlos who voted for R
17:37:43 From Ashen Rana To Harpreet Sahota(privately) : Oh man, that is freaking awesome Ben!
17:38:11 From William Rodriguez : I voted for R :)
17:38:21 From Joe Reis : Funny - Ben’s literally 20 minutes from my house right now. He’s snowboarding at my local spot
17:38:22 From Mikiko Bazeley : Also it’s tough right now for job searching
17:38:48 From Dave Langer : Regarding the poll, ask yourself this. Why would a former C++ OO elitist prefer R over Python? :-D
17:38:58 From Mikiko Bazeley : XD
17:39:00 From Dave Langer : That's me, BTW. :-p
17:39:52 From Ray Givler : I voted for R. Just learning, but I knew I could get some internal R mentors.
17:40:09 From Dave Langer : Wise man, Ray. ;-)
17:40:10 From Monica Royal : Great dedication Ben!! Glad to have you call into the event and hear from you as always! :D
17:40:10 From Joe Reis : Dave - is this you?
17:40:11 From Joe Reis : https://github.com/EnterpriseQualityCoding/FizzBuzzEnterpriseEdition
17:40:27 From Liuna Issagholian To Harpreet Sahota(privately) : sure
17:40:46 From Srivatsan Srinivasan : I have to drop now.. Have a nice holiday time, Merry Christmas and New Year
17:40:54 From Joe Reis : Later Srivatsan
17:40:57 From Eric Sims : See you, @Srivatsan!
17:41:25 From Monica Royal : See you later Sirvatsan!
17:41:57 From Ben Taylor : getting a job isn’t a charity, make sure you sell the value you’ll bring. the employer needs a return on your salary
17:42:01 From Carlos Mercado : relationships != transactions. Great point Jean-Sebastien.
17:42:24 From Mikiko Bazeley : +1 to Ben’s point
17:42:24 From Dave Langer : @Joe - No, need to check this out. Hopefully it's using EJBs!
17:42:30 From Ben Taylor : sorry about my strange joining... hopefully not too distracting with my coin toss silence
17:42:54 From Matt Housley : @ben There was a lot of envy on this call.
17:43:03 From Mikiko Bazeley : Tons lol
17:43:04 From Carlos Mercado : @ Dave what is your list of jobs that are data adjacent but don't call themselves data scientists?
17:43:10 From Joe Reis : Which lift you on @Ben?
17:43:33 From Dave Langer : @Joe - Look at all the GoF.
17:43:47 From Eric Sims : Young ones... because Greg is sooooo old :)
17:44:26 From Greg Coquillo : lmao
17:44:36 From Carlos Mercado : Titles for jobs you can chase after:
Data Scientist
Data Analyst
Quantitative Analyst
Business Analyst
Financial Analyst
Database Analyst

You can also chase after things like Consulting where data science background gives you an in to doing data science work under a different title.
17:45:13 From Carlos Mercado : Economics is the best undergrad clearly lol.
17:45:32 From Ray Givler : Gotta go. Take care folks! See you in the new year!
17:45:36 From Eric Sims : I have also just searched skill terms like "regression" in LinkedIn because then I'll get more data-oriented roles
17:46:04 From Lalita : Thanks a lot everyone for your advice
17:46:04 From Liuna Issagholian : yaaaas Carlos!
17:46:13 From Ashen Rana : DataOps roles maybe?
17:46:24 From Mikiko Bazeley : Women Who Code, Tech Ladies, and AnitaB are also really good places to put resumes or look at jobs
17:46:45 From Carlos Mercado : Eric is ACTUALLY HACKING LinkedIn right now
17:46:46 From Sarah Nabelsi : Two more things I would add here is: 1. Tailor your resume to the job you’re looking for, don’t just dump everything. 2. It’s going to take time, so be kind to yourself :)
17:46:52 From Ameya Dhaygude : Awesome advice Eric
17:47:01 From Carlos Mercado : Eric with the JUMBO brain right now.
17:47:02 From Mikiko Bazeley : With Tech Ladies you can get access to the hiring manager
17:47:06 From Thom Ives : Nice Sarah!
17:47:17 From Carlos Mercado : Eric stop job hunting, you're going to work with me at Guidehouse.
17:48:13 From Akshay Mandke : how are the scores calculated? is that the only metric for building a model?
17:48:25 From Carlos Mercado : I would need a re-statement of the question.
17:49:03 From Carlos Mercado : all transformations have costs in terms of interpretability and the unit of measure for things like error, coefficients, etc. So be careful in general.
17:49:20 From Ben Taylor : listen to Dave... he speaks data gospel!
17:49:28 From Mikiko Bazeley : +1 Dave!
17:49:57 From Akshay Mandke : Always be mindful of the trade offs and standard errors
17:50:19 From Dave Langer : Wait a second, I thought I was a heretic?
17:50:28 From Dave Langer : My R-loving ways and all that.
17:50:35 From Mikiko Bazeley : Dave - Should have ended with “All done in R"
17:50:45 From Dave Langer : ^
17:50:52 From Mikiko Bazeley : Always Be Closing
17:50:57 From Kamarin Lee : ^
17:51:06 From Carlos Mercado : ABC ^
17:52:00 From Eric Sims : "Not trying to stump" - Famous last words immediately before stumping ;)
17:52:05 From Mikiko Bazeley : @Dave - If it helps, my reason for sticking with python is becuase it’s the last language I needed to use extensively and I didn’t feel like context switching
17:52:09 From Mikiko Bazeley : So basically laziness
17:52:23 From Eric Sims : Googling now...
17:52:39 From Joe Reis : Greg - hit me up. I’ve got someone you should talk with about federated learning
17:52:41 From Dave Langer : Not to worry, I've coded in so many languages over my career. My love with R will change before I retire I'm sure.
17:53:33 From Dave Langer : @Mikiko - Totally makes sense. If you're awesome at Python, stick with it.
17:53:38 From Eric Sims : Greg's questions are top-shelf!
17:53:55 From Mikiko Bazeley : ^
17:53:57 From Greg Coquillo : Got you Joe
17:54:05 From Carlos Mercado : Keywords for Federated Learning:

  • Transfer Learning
  • Edge Computing
  • Blockchain AI

Look into these keywords to fully get the scope of Federated Learning.
17:54:06 From Greg Coquillo : Thanks! Eric!
17:54:43 From DavidTello : @Lolita, I was in the same situation last year. Here are a few points of advise. (1) Looking for a job is hard and takes time. I was in the job market from Sept 2019 to March 2020. I had several hundred rejections and only was able to get 5 interviews all together. (2) Harpreet once shared that he got an interview from emailing a manager and share the format of his email. After, I got a rejection for a position with the Army, I applied online. I went ahead an email the hiring manager using Harpeet’s style of email. Within a few days a got a direct call from the hiring manager and she asked me to come in for an interview. Sometimes it is well worth it to take a chance. (3) Be more than willing to take a job at a small town in the middle of nowhere. When the Bank where I worked today interviewed me, they asked me: “Are you willing to move to Topeka, KS. A small town of about 100K people.
17:57:05 From Lalita : Thanks @DavidTello for the advice. Would you mind sharing the template of that email if you have it. Would really appreciate it
17:57:05 From DavidTello : @Lolita, The Bank literally gave my family everything we asked for including relocation money. I started there too working there two weeks before the pandemic. 9 months forward, I still working for the Bank, but I now do it from a home office in Phoenix, AZ where I live and I can see me doing it for at least 6 more months as everyone in the Bank is in WFH mode.
17:57:30 From Carlos Mercado : This also gets deep into the actual hardware itself. The new Apple chip designed specifically for certain ML applications, optimized for certain types of data transfer, etc. I don't know this stuff tbh.
17:57:43 From Kamarin Lee : i was just thinking about that^
17:57:58 From DavidTello : @Lolita, please send me an email reminder to [email protected] and give a little time to look for it
17:58:49 From Joe Reis : @greg - https://coral.ai/products/
17:58:58 From Joe Reis : Have fun with those
17:59:00 From Sarah Nabelsi : Fun discussion guys! I have to hop off, but lovely seeing everyone! Happy holidays. Ben, super jealous—looks amazing!
17:59:13 From Greg Coquillo : thanks Joe!
17:59:33 From Monica Royal : Go Ben!!
17:59:41 From Joe Reis : Ben - did your kid biff it?
17:59:49 From Joe Reis : Saw that
17:59:53 From Lalita : @DavidTello thank you so much for the help. I will share it for sure
17:59:56 From Dave Langer : If I was 1/2 as cool as Ben I would be 5x cooler than I am now.
18:00:05 From Mikiko Bazeley : ^
18:00:10 From Joe Reis : Dammit, wish I was up there right now
18:00:10 From Monica Royal : That looks so fun!

18:00:19 From Ashen Rana : ^ yes !
18:00:37 From Akshay Mandke : I’m in Canada I need I should pull up a Ben
18:01:02 From Mikiko Bazeley : I’m low-key envious thet anyone who looks at this recording is going to see Ben snowboarding
18:01:11 From Jacqueline Lefèvre López : I have to go, but this was awesome. I learned a lot tonight!
Happy holidays everyone! :)
18:01:40 From Carlos Mercado : Oh you're gonna get Ben's whole HIPAA rant, it's a good one though!
18:02:11 From DavidTello : @Ben can you please share it with me
18:02:13 From Mikiko Bazeley : I can get behind it, it was why we could offer social community at Livongo/Teladoc even for people to engage with their own data
18:02:18 From Mikiko Bazeley : *couln't
18:02:21 From Mikiko Bazeley : HIPAA
18:02:47 From Mikiko Bazeley : (^ in response to Ben’s HIPAA rant)
18:02:53 From Ashit Debdas : I got to go. thank you .. have a happy weekend
18:03:36 From Mikiko Bazeley : +1 Carlos
18:03:49 From Joe Reis : @Carlos - that’s almost buzzword bingo ;)
18:03:50 From Carlos Mercado : @ Greg, Also sent these on LinkedIn. I'm doing a presentation on these papers. It includes federated learning with blockchain notes. Hopefully I didn't butcher anything, I am barely ahead of you on this LOL.
18:04:06 From Carlos Mercado : you call it buzzword bingo, I call it technical sales XD XD
18:04:18 From Mikiko Bazeley : Tomatoe, Tomato
18:04:20 From Joe Reis : basically
18:04:35 From Greg Coquillo : lol
18:04:40 From Greg Coquillo : thanks so much Carlos!
18:04:41 From Joe Reis : Cloud Blockchain for Edge ai
18:04:49 From Mikiko Bazeley : BAM! Bingo!
18:04:54 From Greg Coquillo : I like that!
18:05:09 From Carlos Mercado : Joe, that's the title of my next white paper, I'll tag you when its public.
18:05:22 From Joe Reis : deal
18:05:46 From Eric Sims To Harpreet Sahota(privately) : Hey Harpreet, I know Matthew Blasa announced on LinkedIn a couple of days ago that he just got a new job with Brinks. I thought he might be interested in sharing the news, plus it could be motivating for your listeners. Just dropping it on your radar.
18:06:00 From Harpreet Sahota To Eric Sims(privately) : AWESOME! Is he here still?
18:06:11 From Eric Sims To Harpreet Sahota(privately) : Yes! He just doesn't have his camera on
18:06:32 From Eric Sims To Harpreet Sahota(privately) : He just got a job as a Data Governance Analyst with Brinks
18:07:23 From Lalita : I learnt Python using Kdnuggets and medium articles. They are really good and descriptive.
18:07:27 From Mikiko Bazeley : Well if you learn python you’ll be learning the 1st or 2nd most popular language for the group depending on who you talk to, so there’s that
18:09:09 From Mikiko Bazeley : Maybe see if you can convince some of the leads to foot the budget for workshops and training?
18:09:17 From Carlos Mercado : I am gonna drop guys; getting late for free. Happy Holidays all. Stay safe & sane. If you're looking for a new game, Unstable Unicorns has become my favorite.

Good luck to all people looking for jobs too!
18:09:28 From Joe Reis : Later Carlos. Looking forward to your paper!
18:09:30 From Mikiko Bazeley : Bye Carlos!
18:09:39 From Eric Sims : Later Carlos!
18:09:43 From Monica Royal : See you later Carlos!
18:09:45 From Ashen Rana : See ya Carlos !
18:10:19 From Naresh Reddy : Congrats Matthew
18:10:40 From Lalita : congratulations 🥳
18:10:59 From Mikiko Bazeley : There is never
18:11:06 From Mikiko Bazeley : Kidding but kind of not)
18:11:23 From Kamarin Lee : congrats!
18:11:33 From Harika Panuganty : Congrats Matthew!
18:11:42 From Mikiko Bazeley : Congrats!!!
18:11:52 From Kamarin Lee : be right back!
18:11:54 From Ashen Rana : Congrats! :)
18:12:40 From Ameya Dhaygude : Thank you Harpreet, Dave, Joe, and Matt for advice on SAS to Azure
18:12:54 From Eric Sims : 🥳 🥳 🥳
18:13:03 From Liuna Issagholian : Congrats Matthew
18:13:15 From Joe Reis : NP @ameya
18:13:19 From Naresh Reddy : I have question Harpreet
18:13:37 From DavidTello : I have one.
18:13:59 From Lalita : Thanks everyone. It was nice interacting and getting great advices from everyone
18:14:12 From DavidTello : In the past week or so, I been feeling what I think could be
18:14:16 From Matthew Blasa : Thank you everyone
18:15:02 From DavidTello : “survivor’s guilt” because of everything that is going on. Is anyone else experiencing something similar
18:15:47 From Dave Langer : @Naresh - It is often the amount/level of software engineering requires.
18:15:53 From Joe Reis : “Data Scientist” as a title has turned into a honeypot for some companies who want to attract more candidates. What those companies really mean is data analyst
18:15:54 From Dave Langer : *required
18:16:28 From Ben Taylor : thanks for this, too hard multitasking! have a great weekend!!
18:16:31 From Joe Reis : I actually think the role “data scientist” will be split into more accurate roles in the next few years. It’s too overloaded.
18:16:39 From Dave Langer : ^
18:16:45 From Mikiko Bazeley : ^
18:16:54 From Joe Reis : It’s becoming like “Sandwich artist”
18:16:59 From Ashen Rana : ^ Data Scientists will be specialized roles
18:17:02 From Greg Coquillo : Have a great one Ben!!
18:17:09 From Monica Royal : I hope you are right @Joe. It would be easier for all to have specialized roles
18:17:20 From Joe Reis : Analyst is now cool again though
18:17:27 From Joe Reis : For the last few years, analyst was a 4 letter word
18:17:51 From Giovanna Galleno : Agree with you @Joe!
18:20:40 From Matt Housley : Companies will also use the title “data scientist” to promote themselves with investors. I’ve witnessed many cases where a company doesn’t care at all about data, but promotes some people to the data scientist role so they can tell the prospective investors in their next round that they have “AI.”
18:20:50 From Eric Sims : 200 years ago, physicians didn't really have specialties, but as medical science has progressed the roles and titles have become super specialized. It took trial and error and regulation. It's not going to take 200-ish years to subdivide data science, but I think it's good to keep the pattern in mind that this isn't the first time an industry has matured and specialized.
18:22:37 From Eric Sims : I hadn't even thought about negotiating a kibble allowance into my compensation package...
18:22:37 From Greg Coquillo : +1 Vin!!!!!!
18:25:31 From Matt Housley : A friend of mine was hired as a data engineer at a small startup. The founder told him that they needed data engineering so that he could tell the investors that the company “had data.”
18:25:45 From Mikiko Bazeley : I mean he’s not wrong XD
18:25:57 From Matt Housley : Haha, yeah for sure
18:25:59 From Joe Reis : ^ yep. Done many pitch decks for VCs to get the AI/data multiple on the valuation
18:26:04 From Mikiko Bazeley : Investor milestones essentially go from:
18:26:06 From Matt Housley : It’s absolutely part of the startup game now.
18:26:14 From Mikiko Bazeley : Have app => Have data => Have models
18:26:14 From Matt Housley : Or part of selling any company.
18:26:29 From Mikiko Bazeley : +> have USers
18:26:42 From Mikiko Bazeley : It’s a pretty big jump valuation wise
18:27:26 From Albert Bellamy : did research, right after gunning down a squadron of rebel pilots...
18:27:34 From Mikiko Bazeley : XD
18:28:04 From Joe Reis : “Red 5 standing by”
18:28:07 From Thom Ives : Head Rebel Alliance Fight Pilot - ERIC SIMS
18:28:11 From Albert Bellamy : crushing my dreams Eric....
18:28:19 From Harpreet Sahota : Last call for questions!
18:28:38 From Harpreet Sahota : Type them out!
18:28:41 From Eric Sims : Gold Leader, standing by!
18:29:14 From Albert Bellamy : remind me never to hang out with 2013 Dave Langer. he sounds like a real snob.
18:29:29 From Joe Reis : Vintage Langer 2013
18:29:42 From Matt Housley : A total short hair. Could have worked at the CIA.
18:29:50 From Joe Reis : lol
18:29:56 From Eric Sims : Short-haired, non-heretical Dave? Hard to imagine.
18:30:04 From Albert Bellamy : I prefer "Langer Classic"
18:30:08 From Akshay Mandke : I agree. I worked in fraud investigations and sometimes Excel can solve your problem so not every business problem is a data science solution.
18:30:10 From Mikiko Bazeley : XD
18:30:14 From Joe Reis : He had an earring too
18:31:07 From DavidTello : Does any of you thinks that DS position requirement could eliminate the college degree (any degree) requirement at any point in the near future?
18:31:21 From Joe Reis : 2013 Joe would’ve been good friends with 2013 Langer. At the time, ML was the cool thing
18:31:34 From Dave Langer : Dave of 2013 was very passionately wrong about so many thing - including short hair.
18:31:37 From Matt Housley : @DavidTello I think job recs should eliminate this requirement.
18:31:39 From Joe Reis : Then I saw the title morph into…other stuff…
18:31:43 From Vin Vashishta : @David. YES!
18:31:55 From Thom Ives : Proven problem solving rules, BUT many businesses see degrees as a right of passage and an easy first base qualification.
18:32:03 From Eric Sims : ^
18:32:14 From Matt Housley : Academia does a poor job of preparing people to be data scientists. Degree or not, you have to develop a lot of the skills on the job.
18:32:32 From Joe Reis : Colleges have rent seeked their way to being a job filter
18:32:46 From Dave Langer : What @Matt said. Same is true for software engineering.
18:32:54 From Matt Housley : @Thom To your point, getting the first job is really tough without the credentials.
18:33:15 From Joe Reis : The university where Matt and I teach has seen record enrollment in the data programs
18:33:29 From Joe Reis : So, Matt and I will be busy this Spring
18:33:36 From DavidTello : I agree completely. I have started to believe that the ”college degree requirement” will most likely die after the pandemic
18:34:07 From DavidTello : @Joe where you guys teach?
18:34:15 From Joe Reis : University of Utah
18:34:22 From Matt Housley : @DavidTello Seems like this is happening much faster for software engineering.
18:34:37 From Eric Sims : Gotta drop off - my wife and I are going to watch the new Mandalorian episode! #DateNight
18:34:44 From Joe Reis : As a general rule, I think data’s 5-7 years behind software
18:34:59 From Greg Coquillo : Have a great one Eric!
18:35:02 From Giovanna Galleno : bye Eric! :D
18:35:04 From Monica Royal : See you later Eric!
18:35:09 From Thom Ives : Matt - that's correct, but those hiring people that realize that portfolios of proven work are key, will look for that, BUT your portfolio of work, showcased online, better be first class!
18:36:02 From Lalita : see you later everyone
18:36:29 From Naresh Reddy : Thank you Monica, Dave, Giovanna, Joe, Vin, albert, Eric, Mikiko and all the attendees :)
18:36:34 From Joe Reis : Thanks all
18:36:59 From Thom Ives : Great job Harpreet!
18:37:09 From Thom Ives : Harpreet started is because he cares!
18:37:13 From Greg Coquillo : Awesome job Harpeet!!!
18:37:15 From Joe Reis : yep
18:37:18 From Giovanna Galleno : My pleasure! Thank you for your question @Naresh! :D
18:37:49 From Mikiko Bazeley : Harpreet you’e the best!
18:38:01 From Mikiko Bazeley : Virtual pub
18:38:16 From Ameya Dhaygude : You are awesome Harpreet! Thank you.
18:38:17 From Giovanna Galleno : Agree with Mikiko! :D
18:38:20 From Matt Housley : Thanks Harpreet!
18:38:28 From Monica Royal : This is so awesome, I love this event and you are helping so many! :D
18:38:35 From Akshay Mandke : pd.append(“refill”)
18:39:16 From Naresh Reddy : There is so much learning here and at o cost
18:39:20 From Juan Francisco : Happy Holidays
18:39:22 From Naresh Reddy : 0

18:39:28 From Monica Royal : Happy Holidays all
18:39:29 From Mikiko Bazeley : Happy holidays everyone!
18:39:32 From Ashen Rana : Thank you all for your company. And thank you as always, Harpreet. Happy holidays, Happy New year and stay safe everyone! <3
18:39:33 From Ameya Dhaygude : Happy Holidays Everyone. See you on Jan 8th.
18:39:36 From Liuna Issagholian : Thanks Harpreet. Happy Holidays everyone :)
18:39:37 From Thom Ives : God Bless Everyone!
18:39:38 From Akshay Mandke : Happy Holidays everyone. So glad to join this today.
18:39:46 From William Rodriguez : Happy holidays and thank you for letting me listen in on the conversation, very interesting and useful to hear from people who are in the field!
18:39:47 From Naresh Reddy : Happy holidays and New years !
18:39:49 From Harika Panuganty : Happy Holidays everyone! See you all in the new year. Really enjoyed happy hour today!

Article Comments