Data Science Stack 2021 // This Week in Data - Week 19
My Data Science stack going into 2021 plus funding news, weekly reading list, and more.
Welcome back to Dataset Daily! Here’s an outline for the coming week…
💻 Sample Code Wednesday: Python Workflow Orchestration
🏭 Industry Thursday: Crypto/DeFi Analytics for Bitcoin 30k
📈 Market Friday: Updates to the Portfolio
Thanks for reading Dataset Daily! Your support helps us keep improving and offering this newsletter 4-5 times each week.
We’ve decided to permanently enact the holiday discount and offer subscription to the newsletter at $9/mo.
🎉 Week 19 - Kicking off the New Year 🎉
Let’s kick off the New Year right. Here’s to an awesome year of data. Personally, I’m committing to writing more content and meeting as many positive-minded innovators as possible.
Agenda: Funding & Acquisitions, Book List, 2021 Tech Stack
📰 Quick News 📰
📈 Funding & Acquisitions 📈
RingCentral acquired DeepAffects for its speech intelligence. DeepAffects builds AI to find insights from business meetings while RingCentral sells conferencing, phone, and messaging services. Likely looking to put DeepAffects right on top of their existing products.
Nuro acquired Ike autonomous trucking. Nuro is an autonomous delivery company while Ike was focused on self-driving trucks. Plenty of synergies here, including existing tech partnerships.
Octopus raises $200M for its renewable energy and data-driven load balancing. Octopus is a renewables focused energy company. A big part of their innovation comes from their growing Data Science team. In 2018 they acquired a Data Science team dedicated to predicting loads and whatever else is important to an energy company. Cool stuff.
WeRide raises $200M for autonomous driving in China. Another mobility company pushing towards L4. Supposedly WeRide launched a robotaxi service in Guangzhou in 2019. I’m excited to look back a decade or two from now and see which self-driving companies survive.
📡 Cool finds & releases 📡
🎧 How Bad Is Your Spotify - Fun project from Pudding that connects to your Spotify and (harshly) judges your music taste. Really fun application and one of my favs from 2020 - lots of inspiration for other things we can build from connecting to services people use every day.
📚 What I'm Reading 📚
Last week: didn’t read anything, was busy traveling, relaxing, and not being busy.
My Book list for Q1 of 2021: Machine Learning Design Patterns (AI/ML), Naked Statistics: Stripping the Dread From the Data (Stats), Little Fires Everywhere (Fiction), The Welcoming (Fiction)
⌨ 2021 Tech Stack ⌨
Languages: Python, Go
Python remains my go-to language for 2021. Been continuing to level up my Go skills as it continues to grow in popularity amongst my backend friends. I almost never write production Go but love the language regardless. I ship code wildly fast with Python and it’s the Data Science community’s go-to language so it remains my go-to language. I use VS Code as my editor.
Database: PostgreSQL, Firestore
I almost feel guilty sometimes admitting I love Firestore and NoSQL. A lot of database stuff I do remains PostgreSQL by default, but these days any time I’m trying to quickly ship something I go with Firestore for its ease of use and real-time friendly features. A few different fun ones I’ve been trying out but am a long way from adopting, i.e. RethinkDB.
Machine Learning: TensorFlow, Scikit-Learn
Started as a TensorFlow guy for my Deep Learning back in college and have no plans on switching or learning something new at the moment. I still use Scikit-Learn for basically everything else. It’s so entrenched in my workflow that switching costs are way too high for now.
Collaborative Data Science: Deepnote
Deepnote is one of my favorite products from 2020 and their team is as awesome as their product. Definitely recommend checking out Deepnote if you collaborate with others on your notebooks.
Visualizations & Dashboarding: Dash/Plotly, Tableau, Streamlit
Love me some Dash/Plotly. Started the year using Streamlit constantly but transitioned to spending most of my viz time with Dash/Plotly. Every once in a while I collaborate on BI stuff with Tableau. If it’s BI I go with Tableau, if it’s more prediction or computation heavy it lives in Plotly or Streamlit.
Data Orchestration: Prefect
I put off getting truly comfortable with Airflow for the longest. Was able to pick up Prefect pretty easily with their examples and docs. If you’ve been putting off orchestration I recommend checking out Prefect.
Others: Airtable, Notion, Scrivener
I love Airtable. I use it constantly to quickly throw together data. I use it as a database sometimes with its Python wrapper for MVPs, and I use it to embed spreadsheets on my websites.
Continuing to use Notion for organizing my life. Still haven’t fallen prey to the Roam Research wave - maybe soon.
I use Scrivener for managing book manuscripts. When I started writing I googled best book writing apps and Scrivener was widely regarded as the best. Been good to me so far.
💼 Job Board 💼
No fresh job listings popped up over the holiday.
(Contact us to get your job listing posted here)
Enjoy Dataset Daily and want access to our Wednesday through Friday posts? You’ll get sample code, industry deep dives, and our weekly market review. Join now for $9/mo.
💻 Sample Code Wednesday
🏭 Industry Thursday
📈 Market Review Friday
Thanks for joining us for another week of our community newsletter!
See you in the new year!
About the writer: Luke Posey is the founder and lead writer of Dataset Daily. He is passionate about building and investing in great data-driven products. He’s the co-founder of Spawner.ai, an analytics company building data tools for traders and investors. He started his career as a Machine Learning Engineer after studying Electrical Engineering. He currently resides in St. Louis, Missouri.
See you next time.