🚨 I recently completed 10 years in the IT industry. 🎉
Looking back, I realize every phase of my career taught me something valuable.
The first 3 years were the toughest.
I was always under pressure, constantly learning, and sometimes working with the fear of making mistakes.
But today I realize — those were the years where I learned the most.
The biggest lesson from that phase:
Never hesitate to ask questions.
If you don't understand something, ask. There is no growth without curiosity.
After the first few years, I learned that surviving is not enough. You need to start thinking beyond your assigned tasks.
Don't just complete the work given to you.
Ask yourself:
➡️ How can I improve this process?
➡️ How can I solve a bigger problem?
➡️ How can I create more value for my team?
With 10 years of experience, I have learned one important thing:
Growth doesn't always come from getting a new assignment.
Many times, you can create new opportunities within your current work.
Build a prototype.
Suggest improvements.
Automate repetitive tasks.
Solve problems that nobody asked you to solve.
That's how you become valuable.
The IT industry rewards people who don't just execute tasks, but those who think, create, and take ownership.
10 years later, my biggest learning is:
Don't wait for opportunities. Create them. 🚀
What is one lesson your career has taught you?
#DataEngineering #SoftwareEngineering #CareerGrowth #Tech #Leadership #Learning
If you use AI chat interfaces regularly, the experience matters as much as the model.
Claude’s web chat gets a lot of the small things right, especially when you’re reading documentation. The answers feel easier to navigate, and in my experience, Opus gives me more substance with less noise.
ChatGPT still has work to do on the reading experience. The difference isn’t only the model; it’s how information is organized and presented.
The best AI won’t just give you an answer. It’ll make that answer easier to use.
Which interface works better for you?
Your company builds an AI agent.
It can query databases. ✅
It can access business tools. ✅
It can execute operations. ✅
But what happens when someone tricks it into performing an unauthorized action? 😳
Databricks is introducing security screening for AI-agent tool calls.
Should AI agents have the same security restrictions as human employees?
Your company has 10,000 PDF documents.
Management wants an AI chatbot that answers questions from them.
But first, someone needs to prepare the documents.
BigQuery now supports AI-powered document chunking directly through SQL! 🤯
Is SQL becoming an important language for AI engineering?
Me: “Why is this customer 126 years old?”
Source team: “We put 1900 when we don’t know the birth year.”
Me: “So our most loyal customers are actually missing values.” 🫠
Manager: “Can you make the query faster?”
Me: “Done. From 20 minutes to 10 seconds.”
Manager: “Great. Can we add these 14 joins?”
The query has requested its old timing back. 😭