Checking nulls, duplicates, row counts ...is kinda of straightforward....
But what about actual business rules!
Like revenue should be calculated in a certain way, some records should be excluded, dates should follow some rule and more....
Do you guys keep these as separate tests or validate them as part of the ETL itself or any....
Curious how people do actually handle this in real projects.
Not talking about some crazy DB failure or anything.
Could be something silly like a missing WHERE, wrong JOIN, duplicate rows, bad UPDATE....
Like a small mistake that looked harmless but ended up causing a big headache ....🤕
Curious what ppl have seen in real projects.
Do you rollback everything and start again, or resume from the failed step?
Also how do you avoid duplicate records when the job is rerun?
Curious how ppl handle this in real prod pipelines....
Is there any Tableau governance capability to identify things like:---
Similar or duplicate workbooks
Unused workbooks or views
Failed or frequently failing extract refreshes
Similar or duplicate data sources
Unused data sources/dashboards
Workbook/dashboard usage and adoption
Dependencies between workbooks, views, and data sources
I’m curious to know what Tableau provides natively for these governance-related insights through Tableau Server/Cloud, Admin Insights, Metadata API, or other APIs.
How are these governance use cases usually handled in Tableau?...
I was thinking about this while working with ETL and data validation. Suppose a pipeline is processing 10 million records and around 500 records have some data quality issues.
Should the whole pipeline fail, or should it continue loading and just raise an alert? or
should the bad records be separated and the good records continue?
Iam curious how people usually handle this in real production ETL pipelines, especially when a small data issue can block the entire downstream process.
I have a small confusion with Azure Data Factory. Pipeline, Dataset, and Linked Service are used together, but I want to understand how exactly we should differentiate them in real projects.
For example, if we are moving data from Azure SQL to ADLS, I understand Linked Service is mainly for the connection, Dataset is for the data we are accessing, and Pipeline is for the actual workflow.
Is this understanding correct! And in real projects, how do you decide when to create a new Dataset or Linked Service and when to reuse the existing one?...
Imagine you’re starting your software career from zero again in 2026.
With AI changing development, the job market being competitive, and new technologies coming up every day, what would you actually choose to learn?
Would you focus on DSA, backend, cloud, AI/ML, data engineering, cybersecurity, or something else?
And more importantly, what would you avoid wasting time on this time around?
For me, the biggest benefit of ChatGPT isn’t getting a complete answer. It’s getting past the blank page.
Whether it’s coding, writing, learning something new, or planning something, having a starting point makes everything easier.
I’m curious if anyone else uses it more as a "get me started" tool than an answer machine.
Guys, what is one thing you wish you knew earlier about ChatGPT?
I’ve been using ChatGPT for almost everything... 😂 — learning, work, random doubts, and sometimes just asking stupid questions.
What’s one feature, prompt, or trick you found that actually made ChatGPT more useful for you?
I recently came across Databricks AI/BI and was curious to know how people are finding it.
It looks like Databricks is trying to bring BI and analytics more directly into the Databricks platform, with dashboards and Genie for asking questions in natural language.
Has anyone actually tried AI/BI in a real project?
How is it compared to Power BI or Tableau from your experience? Is it good enough for regular BI use cases, or is it still better to use a separate BI tool?
Would like to know your experience, especially if you have used both.
Hey guys,
I’m planning to learn Java for backend development. I know some basic programming, but I’m new to Java.
I’m confused about where exactly to start. Should I first complete Core Java and then go for Spring/Spring Boot?
Like Java basics ->OOP -> Collections-> Exceptions -> Maven ->Spring ->Spring Boot -> REST API -> JDBC/JPA?
Or is there any better order to learn these?
Also, if you guys have an idea for learning Java and Spring Boot fast
I don’t want to just watch tutorials. I want to learn properly and build something while learning.
What would you suggest?
I recently came across Databricks AI/BI and was curious to know how people are finding it.
It looks like Databricks is trying to bring BI and analytics more directly into the Databricks platform, with dashboards and Genie for asking questions in natural language.
Has anyone actually tried AI/BI in a real project?
How is it compared to Power BI or Tableau from your experience? Is it good enough for regular BI use cases, or is it still better to use a separate BI tool?
Would like to know your experience, especially if you have used both.