El ecosistema de ingeniería de datos evoluciona a altísima velocidad. Ya es tiempo de que subas de level y conozcas más allá de numpy, pandas y matplotlib.
DuckDB nos permite hacer OLAP desde nuestro navegador web y tener un motor que funciona bastante bien con Parquet. MotherDuck motherduck.com está buscando ofrecer como Saas DuckDB a gran escala.
Polars 🐻 : pola.rs
Polars permite como Pandas procesar dataframes pero a velocidades superiores y sin depender 100% de que los datos quepan en memoria.
Pydantic: docs.pydantic.dev
Estas herramientas nos permiten tener las características de static typing y robustecer nuestros pipelines. Así como cuando en Java tienes que declarar los tipos, estos sería el equivalente en python
What is the difference between a Data Engineer and a Data Architect?
🧵[1/x]
A data engineer looks at the immediate set of requirements and works towards that. In other words, data engineers build, rebuild, and tear down. ⚒
Need a new field in the report? Let's just build the whole thing. ⚒
🧵[2/x]
Data Architects think ahead in terms of capacity planning. X years from now, Y will happen, so we'll need to consider Z. In other words, Data Architects look at the full requirements and build it once.😎
This means less waste of money for the company in the long run.
"Blessing and misfortune are two sides of the same coin. One extreme can transform into another, and there is no right or wrong to this."
🧵[1/x]
During the Han Dynasty, an old man (Sai Weng) living on China’s border one day lost his horse. His neighbors all said what terrible luck that was, and sympathized with the old man. But Sai Weng said: “Maybe losing my horse is not a bad thing after all.”
🧵[2/x]
Lo and behold, the next day the old man’s horse returned, together with a beautiful female horse alongside him. All the neighbors exclaimed: “What great luck!” But the old man responded: “Maybe this is not such good luck after all.”
Some #GoogleCloud professional certificates on Coursera have off-platform certification exams. For a limited time, you can get a discount voucher for 20% off the cost of the exam.
🚨⚠️People issues are the biggest risk to funded startups.
55% of startups fail because of people problems, according to a study by Harvard, Stanford, and University of Chicago researchers.
🧵[2/x]
1. Minimize unnecessary micromanagement
Micromanaging can be helpful in certain situations, the most effective leaders aim to delegate work in order to scale both themselves and their businesses. Our data suggests that micromanaging can be a fatal flaw for CEOs.
Tracking your Uber Rides and Uber Eats expenses through a data engineering process
Technologies and skills:
Python, Docker, Apache Airflow, AWS Redshift, Power BI, data modelling, Task schedulling, ETL and ELT processes, Data warehousing, Cloud