Datacast

Datacast

Download on the App Store

Datacast episodes

  • Episode 9: Diving Into Data Engineering with Mark Sellors

    Show Notes:



    • (2:05) Mark revisited his brief stint during college back in the 90s.

    • (3:38) Mark’s first job as a Business Consultant at Dixon Stores Group was a humbling experience.

    • (5:41) Mark stressed the importance of customer empathy that lasts with him throughout his career.

    • (6:45) Mark talked about his next long-term job as a Configuration & Support Engineer at Orange PCS, where he built solid technical skills in IT and data work.

    • (12:51) Mark discussed his next gig working as a Principal Specialist at T-Systems and then doing freelance work in IT and data analytics.

    • (16:23) Mark reflected on the sabbatical years he took a break from working.

    • (19:10) Mark shared an overview about his current employer, Mango Solutions.

    • (20:34) Mark discussed his first big project working at Mango as a senior IT consultant.

    • (25:57) Mark then transitioned into a technical architect role, where he became the bridge between the data science and the IT worlds.

    • (28:33) In reference to his talk “An operating model for R”, Mark stressed the importance of policy, procedure, people and policing to build an effective operating model that connects data scientists and IT specialists.

    • (34:32) Mark gave a client use case that his Data Engineering team has been involved with at Mango.

    • (36:48) Mark talked about the cultural challenges of deploying code into production within an organization.

    • (39:35) Mark talked about his key accomplishments in his leadership role as Head of Data Engineering.

    • (43:27) In reference to his talk “R is production safe”, Mark discussed the existing challenges in the R-community to write production code.

    • (50:39) In reference to his book “Field Guide to the R Ecosystem”, Mark shared the key developments in the R ecosystem in 2019 that he’s most excited about.

    • (53:27) Mark gave thoughts on the rise of cloud-based processing technologies for data engineering’s best practices.

    • (56:42) Mark mentioned the Google Cloud Certified Data Engineer Exam that people can take to learn about data engineering.

    • (59:35) Mark emphasized the importance of communication skills to become an organizational leader.

    • (01:01:13) Mark shared his view on the data science ecosystem in the UK.

    • (01:02:18) Closing segments.


    His Contact Info:



    • Website

    • Twitter

    • GitHub

    • LinkedIn


    His Recommended Resources:



    • rayshader R package

    • TensorFlow for R

    • sparklyr R interface for Apache Spark

    • Amazon S3 for Data Storage

    • Google Cloud Platform Podcast

    • Hudl

    • RStudio

    • R for Data Science by Hadley Wickham and Garrett Grolemund



    This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit datacast.substack.com/subscribe
    1 hr 8 min
  • Episode 8: From Underwater Communication to Data Science with Chintan Shah

    Show Notes:



    • (2:07) Chintan talked about his undergraduate education in India at the Dharmsinh Desai University, studying Electronics & Communication.

    • (2:53) Chintan recalled the most useful knowledge from his undergraduate experience.

    • (3:42) Chintan explained his decision to pursue a Master degree in Communication and Signal Processing at Newcastle University in the UK.

    • (5:09) Chintan discussed advanced communication systems knowledge he learned during his Master.

    • (6:25) Chintan talked about his decision to continue his academic journey with a Ph.D. in Underwater Acoustic Communication at Newcastle University.

    • (8:15) Chintan went in depth to explain the study of Underwater Acoustic Communication and how it is different from normal communication systems.

    • (11:45) Chintan discussed his Ph.D. thesis on applying Bayesian theorem for Turbo coding to solve signal processing problems under the water.

    • (15:00) Chintan elaborated on the benefits of doing a Ph.D. degree.

    • (19:01) Chintan mentioned the importance of collaboration in the academic community.

    • (20:18) Chintan talked about his Post-Doc experience at University of Illinois Urbana Champaign.

    • (22:00) Chintan discussed his next career step moving back to the UK working as a digital communication engineer at Schlumberger, a big company in the oil and gas industry.

    • (27:30) Chintan characterize datasets that are particular to oil and gas.

    • (28:58) Chintan examined the technique of vibration analysis in his work.

    • (31:23) Chintan next landed a role as a Software Engineer at a company called SmartFocus in London.

    • (34:52) Chintan mentioned the best resources he used to learn machine learning and data science.

    • (36:51) Chintan recalled his experience searching for a data science job.

    • (38:02) Chintan briefly went over the interesting projects that he has involved with at Avanade, his current employer.

    • (40:34) Chintan talked about his recent transition from a Senior Data Scientist role to a Manager in Advanced Analytics role.

    • (43:35) Chintan shared his view about the data science community in London.

    • (44:44) Closing segments.


    His Contact Info:



    • Twitter

    • GitHub

    • LinkedIn


    His Recommended Resources:



    • Andrew Ng’s Machine Learning Coursera Course

    • MIT’s The Analytics Edge edX Course

    • London’s Data Science Festival

    • Nate Silver’s “The Signal and The Noise”

    • Charles Wheelan’s “Naked Statistics”

    • Philip Tetlock’s “Superforecasting”



    This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit datacast.substack.com/subscribe
    50 min
  • Episode 7: Building Open-Source R packages with Thomas Lin Pedersen

    Show Notes:



    • (2:16) Thomas talked about the study of Food Science and Technology in which he focused on microbiology.

    • (3:15) Thomas stressed the importance of user empathy, something useful he gained from his degree.

    • (4:39) Thomas discussed the reason to pursue a Ph.D. in Bioinformatics at the Technical University of Denmark.

    • (6:10) Thomas talked in-depth about the tools he developed for his Ph.D. thesis, which are able to handle large-scale pangenome analyses using sequential data.

    • (9:11) Thomas talked about using the ggplot2 package for his R package “Find My Friends.”

    • (11:11) Thomas worked on the ggforce package, which aims at providing missing functionalities to ggplot2 during his internship at RStudio.

    • (13:34) Thomas recalled the best learning he got from his internship with RStudio.

    • (15:08) Thomas gave advice for people who want to contribute to open-source projects.

    • (18:57) Thomas shared the experience working on the package ggraph, also known as the grammar of graphics for relational data.

    • (22:02) Thomas discussed 2 other packages, tidygraph and particles, that he built to bring graph and network data into the tidyverse, the very popular collection of R packages designed for data science.

    • (25:37) Thomas provided resources for R users who want to learn more about network analysis and network visualization.

    • (27:05) Thomas went over his job as a data scientist at SKAT, where he handled all the advanced analytics going on in the Danish Tax Authorities.

    • (32:15) Thomas talked about the intuition behind working on patchwork, a package that can combine multiple ggplots in the same graphics.

    • (35:50) Thomas summarized his most recent projects, gganimate (a package that extends ggplot2 to include the description of animation) and tweenr (a package for interpolating data mainly for animations).

    • (40:47) Thomas discussed his current job as a software engineer at RStudio.

    • (43:04) Thomas gave his two cent on the Python and R comparison.

    • (45:53) Thomas talked about using Twitter to share his work, where he has more than 10,000 followers.

    • (47:07) Thomas went over something he works on during his spare time, generative art visualization (check his Instagram account!).

    • (49:09) Thomas gave some thoughts on the tech community in Copenhagen.

    • (51:09) Closing segments.


    His Contact Info:



    • Website

    • Twitter

    • GitHub

    • LinkedIn


    His Recommended Resources:



    • Hilary Parker from Stitch Fix

    • David Robinson from DataCamp

    • Unflattening Comic Book

    • DataCamp's Network Analysis in R

    • Katya Ognyanova's Network Analysis and Visualization with R and igraph



    This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit datacast.substack.com/subscribe
    59 min
  • Episode 6: Data Science in the Travel Industry with Ewan Nicolson

    Show Notes:



    • (2:00) Ewan recalled his undergraduate days at the University of Edinburgh studying Physics.

    • (2:43) Ewan talked about his first job out of school as a Seismic Acquisition Specialist at ION Geophysical.

    • (5:28) Ewan briefly skimmed through his next job as a Senior Research Executive in Marketing Science at BBC.

    • (7:40) Ewan gave some key insights on the intersection between marketing and data.

    • (10:15) Ewan introduced Skyscanner and the interesting data problem that he is helping to solve.

    • (12:39) Ewan went over a couple of major data science use cases at Skyscanner.

    • (18:05) Ewan discussed how Skyscanner builds its travel recommendation system.

    • (23:40) Ewan emphasized the importance of doing A/B testing experiments at Skyscanner.

    • (27:58) Ewan talked about the differences in doing experimentation on both Skyscanner's web app and the mobile app.

    • (29:31) Ewan gave a long answer on how a data scientist can effectively communicate the results to others in an organization.

    • (33:07) Ewan gave his two cents on possible ways a data science team can be situated in a company.

    • (36:40) Ewan shared the 3 most important software engineering skills that a data scientist can develop.

    • (38:25) Ewan stressed the importance of learning SQL.

    • (40:25) Ewan mentioned the important signals to look for in a data-centric company.

    • (43:01) Ewan shared more information about the data science and tech ecosystem in Scotland.

    • (45:40) Ewan gave advice for aspiring data scientists who want to succeed during a data science interview.

    • (49:45) Ewan gave advice for business executives who want to bring data talents into their organizations.

    • (51:00) Closing segments.


    His Contact Info:



    • Website

    • Twitter

    • LinkedIn


    His recommended resources:



    • Airbnb Data Science

    • Stitch Fix Data Science

    • Ben Goldacre’s Bad Science



    This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit datacast.substack.com/subscribe
    58 min
  • Episode 5: Applied Statistics in Data Science with Christopher Peters

    Show Notes:



    • (2:05) Chris recalled his Econometrics and Quantitative Economics study during his undergraduate days.

    • (4:14) Chris talked about his research work focus on oil and gas at the Center for Energy Studies at Louisiana State University.

    • (5:53) Chris gave some insights on how data science and econometrics can solve problems in the energy industry.

    • (7:51) Chris talked about his decision to pursue a Masters in Applied Statistics.

    • (9:02) Chris emphasized the importance of learning statistical theory. In particular, survival analysis is very similar to conversion rate analysis.

    • (10:28) Chris discussed his Master’s thesis work in analyzing the churn rate for Treehouse.

    • (12:05) Chris maintained his own website called Statwonk.

    • (13:18) Chris mentioned the popularity of count data on consumer products.

    • (14:40) Chris gave some recommendations for those who want to learn fundamental statistics: Think Stats, Think Bayes, Statistical Inference, and Statistical Intervals.

    • (16:20) Chris recalled how he got a job as the first data scientist at Treehouse.

    • (16:52) Chris described Treehouse CEO Ryan Carson as a “marketing genius.”

    • (18:34) Chris spotted the number one most challenging aspect for a data scientist within a business - translating technical jargon to comprehensible concepts.

    • (20:11) Chris recalled building a company dashboard from scratch for Treehouse.

    • (22:46) Chris shared the background overview of his next employer, Zapier.

    • (24:25) Chris shared his favorite things working as a data scientist as Zapier.

    • (32:07) Chris recalled building an AnamolyBot to keep track of Slack communication between Zapier team members.

    • (36:33) Chris talked about the open-source project he has been working on during his spare time.

    • (39:10) Chris gave advice for people who want to seek remote positions.

    • (42:20) Closing segments.


    His Contact Info:



    • Website

    • Twitter

    • LinkedIn

    • GitHub


    His recommended resources:



    • Shopify’s Cameron-Davidson Pilon

    • Stitch Fix’s Kim Larsen

    • R for Data Science

    • Hadley Wickham

    • John Tukey’s “Exploratory Data Analysis”

    • William Cleveland’s “The Elements of Graphing Data”

    • #rstats

    • #PyData



    This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit datacast.substack.com/subscribe
    50 min
  • Episode 4: AI in Retail with Saurabh Bhatnagar

    Show Notes:



    • (3:20) Saurabh recalled his college experience.

    • (4:25) Saurabh talked about his first role out of school as a software engineer specializing in database at CA Technologies.

    • (8:06) Saurabh landed database consulting roles with different companies.

    • (9:14) Saurabh gave insights on the differences between database systems now and a decade ago.

    • (11:05) Saurabh shared his experience landing a senior data scientist job with Barnes & Noble.

    • (13:15) Saurabh explained major challenges in hiring good data scientists.

    • (16:10) Saurabh discussed his decision to go work for Rent The Runway.

    • (17:57) Saurabh gave insights on the data problems he had worked with at Rent The Runway.

    • (19:43) In reference to his blog post on scaling machine learning at RTR, Saurabh shared knowledge on structuring a data science team.

    • (21:36) Saurabh gave advice for data scientists to incorporate feedback loops into their workflow.

    • (26:00) Saurabh talked about how to give better pitches to business stakeholders.

    • (29:03) Saurabh showed great appreciation for Rent The Runway’s CEO and Founder, Jennifer Hyman.

    • (30:39) In his post Human in AI loop, Saurabh shared the details of building Deep Dress AI, which can show the Instagram photos on Rent The Runway’s product pages and help the customers find fashion inspiration through these high-quality shots.

    • (34:10) Saurabh talked about unique challenges of doing data science in the retail space.

    • (36:18) Saurabh gave a brief glimpse about his stealth-mode startup Virevol.

    • (36:58) Saurabh gave advice for data scientists who want to pursue the entrepreneurial route and start their own ventures.

    • (38:36) In a fun blog post, Saurabh used Generative Adversarial Network to generate dresses photos and use them as training data for the model. He discussed the potential usage of GAN models for fashion.

    • (41:02) Closing segments.


    His Contact Info:



    • Website

    • Twitter

    • LinkedIn


    His recommended resources:



    • Salesforce AI

    • Elucd

    • Nate Silver's "The Signal and The Noise"

    • Deep Learning Textbook by Ian Goodfellow, Yoshua Bengio and Aaron Courville



    This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit datacast.substack.com/subscribe
    45 min
  • Episode 3: How to be a Social Media Data Whiz with Leni Krsova

    Show Notes:



    • (3:08) Leni explains how studying Theatre Critics in college help her develop analytical thinking.

    • (4:00) Leni briefly her first job out of college in brand management at a lifestyle brand called Life is Porno.

    • (5:54) Leni goes over the most interesting classes she took during her Master of Arts in New Media Studies at Charles University.

    • (7:52) Leni talks about her 1-year residency in the social media team of New Media department at Czech TV.

    • (12:02) Leni reflects on the career lesson she got out from her time at Czech TV.

    • (13:47) Leni discusses her first professional role working with data at Socialbakers, a social media marketing company in Prague.

    • (16:35) Leni talks about choosing R as her main programming language to learn data science.

    • (19:43) Leni explains the study of social network analysis.

    • (21:50) Leni discusses the technical aspects of her Master’s thesis about Czech journalists on Twitter.

    • (25:00) Leni goes over her brief stint as a marketing specialist at Zeleznakoule, a gym in Prague and a community of health enthusiasts.

    • (26:57) Leni talks about her next career move to Seznam, the most visited web portal and search engine in the Czech Republic.

    • (29:11) Leni also volunteers as a mentor for Czechitas, a non-profit organization focused on educating girls about programming, data analysis and graphics.

    • (30:17) Leni describes the data science community in Prague.

    • (32:51) Leni talks about her passion for the tech scene in Stockholm.

    • (35:22) Leni talks about her popular Facebook page Dataholka.

    • (36:45) Leni shares some tips for aspiring data scientists who want to improve their social media presence.

    • (39:00) Leni gives recommendations for social media users regarding concerns about data privacy.

    • (41:30) Leni shares her thoughts on the current state of digital journalism.

    • (43:48) Leni discusses her current plan to pursue a Ph.D. degree in the US.

    • (47:52) Closing segment.


    Her Contact Info:



    • Website

    • Twitter

    • Facebook

    • Instagram


    Her recommended resources:



    • Netflix Tech Blog

    • Spotify Labs



    This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit datacast.substack.com/subscribe
    54 min
  • Episode 2: Becoming a Deep Learning Expert with Deep Narain Singh

    Show Notes:



    • (1:58) Deep gives a brief background of his career.

    • (3:42) Deep explains the discipline of civil engineering.

    • (4:24) Deep talks about his first job out of college as a programmer analyst at Cognizant in Pune, India.

    • (5:26) Deep then became a senior software engineer at the investment bank Citigroup in Mumbai, India.

    • (7:30) Deep discusses his next role as a senior associate for the e-commerce company Sapient in Bangalore, India.

    • (8:57) Deep then went to Barclays Investment Bank back in Pune and worked as an analyst, in which he learned a lot about big data technologies.

    • (10:07) Deep talks about the diverse Indian cultures as he had experience working in many different regions in India.

    • (12:22) Deep recounts the story that propelled him to explore study opportunities in the United States.

    • (16:10) Deep gives practical advice to international students who want to pursue a graduate degree in the US.

    • (17:53) Deep talks about the Machine Learning and Deep Learning classes he took while at Galvanize.

    • (19:27) As a graduate student researcher, Deep built an image captioning system for a startup named FoxType, which helps non-native English speaker write better emails. He goes over the challenges in building this system.

    • (24:55) As a deep learning research intern for Netomi.com, Deep built a system with memory using neural network to solve the shortcoming of understanding long-term conversation and context for conversational agent. He discusses the process he went through.

    • (27:59) Deep talks about his machine learning project in which he built a sentiment analysis engine using the 300k reviews from Trip Advisor.

    • (31:33) Deep talks about another data engineering project in which he built an end-to-end real-time mapping of geo-locations in San Francisco for Uber Price Surge.

    • (37:32) Deep recommend the big data tools and technologies that all data scientists should know about.

    • (39:27) Deep shares his job search experience and how he landed a role at Nvidia.

    • (42:02) Deep provides helpful advice to crack the machine learning interview (aka, learning the fundamentals + networking).

    • (45:14) Deep talks about the autonomous vehicle system he’s working on at Nvidia.

    • (47:20) Deep quickly goes over Nvidia’s company culture.

    • (49:32) Deep digs deep into computer vision problems.

    • (51:34) Deep predicts the future applications of deep learning.

    • (53:40) Closing segments.


    His contact info:



    • Twitter

    • LinkedIn

    • GitHub


    His recommended resources:



    • Google Brain

    • Facebook AI Research

    • OpenAI

    • Salesforce Research

    • Andrej Karpathy's Blog

    • Off the Convex Path

    • Nvidia AI Research Blog

    • Google AI Blog

    • Facebook Research Blog

    • Arxiv Sanity Presenter



    This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit datacast.substack.com/subscribe
    59 min
  • Episode 1: Molecular Biologist turned Data Scientist with Jonathan Leslie

    Show Notes:



    • (2:10) Jon reflects on his academic career and explained why he pursued an advanced degree in Biology. He also went over his post-doc fellowship at Cancer Research UK.

    • (4:45) Jon discusses the zebra-fish project he worked on during his post-doc research at King’s College London.

    • (8:13) Jon talks about the developmental biology work he did at the embryology lab at University College London.

    • (10:00) Jon explains why biological data are noisy.

    • (10:58) Jon explains his motivation behind his transition from academia to industry.

    • (15:04) Jon’s advice for Ph.D. students and post-doc researchers who want to make a similar transition.

    • (16:20) Jon talks about his freelance data science project to build a recommendation systems aiming at women struggling with Type 2 diabetes.

    • (21:32) Jon talks about his experience running his data science consultancy business.

    • (25:32) Jon’s advice to improve data storytelling skill.

    • (26:59) Jon presents a high-level overview of his company Pivigo.

    • (31:11) Jon talks about his role as Pivigo’s Head of Data Science.

    • (34:50) Jon goes over the industries that are Pivigo’s clients.

    • (36:30) Jon goes over the skills that data science talents need to develop.

    • (37:53) Jon’s advice to transition from individual contributor to manager.

    • (41:51) Jon mentions the talks he gave recently at O’Reilly Strata Data Conference in London.

    • (46:48) Closing segment.


    His contact info:



    • Twitter

    • LinkedIn

    • GitHub


    His recommended resources:



    • Datacamp

    • David Robinson's "Understanding empirical Bayes estimation"

    • "R for Data Science" by Hadley Wickham and Garrett Grolemund

    • #rstats



    This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit datacast.substack.com/subscribe
    54 min

About Datacast

From the publisher's feed

Datacast follows the narrative journey of data practitioners and researchers to unpack the career lessons they learned along the way. James Le hosts the show.