Brahim Askiou’s Post

I am very bad on reproducing on the spot at interviews. And given the task I choose to arrive at a tool that will be useful for other tasks. I learn when solving a problem. I experiment. You would never employ me expecting to count in my head, not use the cheatsheet.

Data Engineering is what makes everything else work - getting the right data, at the right time, in the right format. When this layer is strong, everything downstream becomes simpler and more reliable. But preparing for Data Engineering interviews doesn’t always reflect that reality. When I was preparing, most of the effort went into stitching things together - SQL from one place, Python from another, Data Modeling and System Design from scattered resources. The challenge wasn’t just learning, but bringing everything together and applying it within the limited time of an interview. That’s where I see a real gap. Recently came across Pipecode AI (https://pipecode.ai/) , and it felt like a more structured approach covering SQL, Python, Data Modeling, and System Design in a way that aligns with how interviews actually work. If you’re preparing for DE roles, this might be worth exploring. #DataEngineer #TechCareers #InterviewPrep #CareerGrowth #Learning

𝐓𝐡𝐞 𝐈𝐧𝐭𝐞𝐫𝐯𝐢𝐞𝐰 𝐓𝐫𝐚𝐩: 𝐒𝐨𝐥𝐯𝐢𝐧𝐠 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐖𝐫𝐨𝐧𝐠 𝐓𝐡𝐢𝐧𝐠 A person spent 48 hours mastering advanced Python libraries. Then got rejected in the first technical round. 𝑯𝒆𝒓𝒆’𝒔 𝒘𝒉𝒂𝒕 𝒉𝒂𝒑𝒑𝒆𝒏𝒆𝒅: The Interviewer gave him a simple business problem: "How would you identify our most loyal customers using this dataset?" The candidate immediately started talking about: - Random Forest Regressions - K-Means Clustering - Complex Predictive Modeling The Interviewer stopped him halfway. "I don't need a model," the Interviewer said. "I need a list. Could you have done this with a simple SQL filter in 30 seconds?" The candidate was silent. He was so busy trying to look like an expert, they forgot to be a problem solver. 𝘛𝘩𝘦 𝘓𝘦𝘴𝘴𝘰𝘯: - The "best" solution isn't the most complex one. It’s the one that provides the most value with the least effort. - Now, before answering any technical prompt, that person asks: "Is there a simpler way to do this?" If the answer is yes, they start there. 𝑪𝒐𝒎𝒑𝒍𝒆𝒙𝒊𝒕𝒚 𝒎𝒊𝒈𝒉𝒕 𝒈𝒆𝒕 𝒚𝒐𝒖 𝒏𝒐𝒕𝒊𝒄𝒆𝒅, 𝒃𝒖𝒕 𝒄𝒍𝒂𝒓𝒊𝒕𝒚 𝒈𝒆𝒕𝒔 𝒚𝒐𝒖 𝒉𝒊𝒓𝒆𝒅. #CareerAdvice #JobSearch #DataScience #Interviews #ProblemSolving

Learning never stops when you're preparing for interviews. But most candidates prepare answers. Top candidates prepare how to think. Part 7 of the Interview Questions series focuses on what interviewers actually test — problem-solving, clarity, and real-world thinking. Sharing a few important questions that can really help you think better and stay confident. Save this before your next interview. You’ll thank yourself later. #DataScience #InterviewPrep #CareerGrowth #MachineLearning #DataAnalytics #TechCareers #Learnbay

One simple question can decide your interview. It’s not about long answers, but clear understanding. These questions help you prepare better. #MLInterviewPrep #DataScienceJobs #LearnML

Days 7–14/17 ✅ This week was about sharpening both technical depth and interview readiness. 🔹 DSA Practice Continued strengthening problem-solving skills with consistent practice, focusing on patterns, optimization, and interview-style thinking. 🔹 Interview Strategies Spent time refining interview approaches — structuring problem-solving communication, breaking down questions clearly, and thinking through solutions the way interviewers expect. 🔹 Completed Reading: Designing Machine Learning Systems — by Chip Huyen This book gave me a broader perspective on how ML systems are designed beyond just models — covering data pipelines, deployment, monitoring, and production challenges. 🔹 Completed Reading: Fluent Python — by Luciano Ramalho This book completely changed the way I think about Python — helping me understand how to write cleaner, more expressive, and truly Pythonic code. Together, these books gave me a new perspective on technology: ➡️ Chip Huyen’s work deepened my understanding of scalable ML systems ➡️ Luciano Ramalho’s insights refined how I approach software craftsmanship in Python Always learning, always evolving. Also — I’m open to recommendations. What books, resources, or topics have shifted your perspective as an engineer? #Day7to14 #DSA #MachineLearning #FluentPython #DesigningMLSystems #InterviewPrep #SoftwareEngineering #LearningJourney

If I had to start Data Analytics again, I would do this 👇 1️⃣ Start with SQL (not Python) Master querying before anything else 2️⃣ Focus on problem-solving, not just syntax Interviews test thinking, not memorization 3️⃣ Work on real-world case studies early Projects > certificates 4️⃣ Learn how to present insights Communication matters as much as analysis 5️⃣ Prepare for interviews from Day 1 Resume + mock interviews make a huge difference Most people delay these things — and that’s where they struggle. If you're starting today, focus on clarity + consistency, not just tools. What would you do differently if you started again? #DataAnalytics #AnalyticsCareers #CareerGrowth #VeritasLearning

Want to instantly signal to a hiring manager that you write production-ready Python? 🛑 Stop looping through your Pandas DataFrames. In live interviews, you’ll often be asked to: 👉 create a new column 👉 apply business logic 👉 transform data How you solve this ONE task tells the interviewer everything about your coding maturity. Here’s the Good → Better → Best hierarchy you need to know: 🔴 .iterrows() — The Beginner Trap What it is: Iterates row by row as (index, Series) The reality: Every iteration creates a new Series → painfully slow at scale Interview verdict: Avoid. Almost always the wrong answer. 🟡 .apply() — The Comfortable Crutch What it is: Applies a function across rows/columns The reality: Still behaves like a Python loop under the hood Cleaner than loops, but not truly optimized Interview verdict: Acceptable only when vectorization isn’t possible (Be ready to justify it) 🟢 Vectorization What it is: Operate on entire columns at once The reality: Powered by NumPy (C-level performance) → massively faster Examples: * df['A'] + df['B'] * np.where(condition, x, y) * np.select([...], [...]) Interview verdict: 👉 This is what strong candidates default to 💡Before coding, ask: “Should I assume this needs to scale to millions of rows?” If yes → skip .apply() and go straight to vectorization. That one question signals: * performance awareness * real-world experience * production mindset 📌 Save this before your next coding round #Python #Pandas #DataScience #DataEngineering #InterviewPrep #CodingInterviews

🚀 Mastering Two Sum, Three Sum & Four Sum — The Real Game is Time Complexity If you’ve ever prepared for coding interviews, you’ve definitely come across Two Sum, Three Sum, and Four Sum problems. At first glance, they look similar — but the real challenge lies in optimizing time complexity. Let’s break it down 👇 🔹 1. Two Sum — O(n) is the goal The brute force approach (nested loops) gives O(n²), but that’s not acceptable in interviews. Using a HashMap, we can reduce it to O(n) by storing complements while traversing the array. 👉 Key Idea: Trade space for time. --- 🔹 2. Three Sum — Sorting + Two Pointers = O(n²) Here, HashMap becomes messy due to duplicate handling. Instead: - Sort the array → O(n log n) - Fix one element - Use two pointers to find remaining pairs 👉 Result: Clean solution with O(n²) time and no duplicates. --- 🔹 3. Four Sum — Extending the pattern to O(n³) Now we: - Fix two elements - Apply two-pointer technique on the remaining array 👉 Time complexity becomes O(n³) — and that’s acceptable given the constraints. --- 💡 Big Learning: Pattern Recognition > Memorization - 2 Sum → Hashing - 3 Sum → Sorting + Two Pointers - 4 Sum → Nested loops + Two Pointers Each step builds on the previous one. Once you understand the pattern, you can even extend this to K-Sum problems. --- 🔥 Pro Tip for Interviews: Always start with brute force, then optimize step-by-step while explaining your thought process. Interviewers care more about how you think than just the final answer. --- #Java #DataStructures #Algorithms #CodingInterview #TimeComplexity #SoftwareEngineering #SDET #Learning #ProblemSolving