Jayant Velichety, MBA, MS

This is How I Used Category-Based Median Imputation to Reduce Distortion and Bias by 5%: While working on a retail analytics project, I noticed several null values in the review rating column. Dropping rows would shrink the dataset, and using the overall mean could be skewed by outliers. So I chose median imputation, which is robust to extreme values. The key insight: Instead of using the overall median, I calculated the median for each product category and filled missing values accordingly. Why it worked: •Preserved category-level patterns •Reduced the effect of outliers •Maintained realistic distributions •Improved reliability of analysis Lesson learned: Data cleaning isn’t just about filling gaps, it’s about reducing bias and preserving the true story your data tells. If you found this useful, Repost it someone in your network might need it. ~Md Zakir Ahmed | LinkedIn

33

1 Comment

💼 Yesterday, I interviewed at Tiger Analytics — and honestly, most of the questions felt familiar. Turns out, these are the same concepts that pop up in almost every Data Quality engineering interview I’ve seen. That’s why I’ve put together a detailed PDF guide that covers: 📊 Most asked interview questions 🧩 Core data engineer concepts 🧠 Real-world case scenarios 💡 And practical tips to boost your confidence Whether you’re preparing for your first Data engineer role or aiming to level up your interview performance, this guide will help you focus on what truly matters — understanding the logic behind data problems, not just memorizing answers. 📥 Download the PDF, go through the insights, and strengthen your preparation for your next big opportunity! Do share and repost for your reference/network 🚀 Consistency > Cramming. Keep learning, keep iterating. Follow @Dakshitha Reddy for more content #DataScience #DataScientistInterview #InterviewPreparation #MachineLearning #ArtificialIntelligence #CareerGrowth #DataAnalytics #JobPreparation #LinkedInLearning #DataDriven

148

4 Comments

🎓 Day 11: Mastering Table Calculations in Tableau 🧠 What Are Table Calculations? Table Calculations are computations applied after the data has been aggregated and displayed in the view. They operate on the result of the visualization, not the raw data. 📌 Examples: 1. % of total sales 2. Running totals 3. Year-over-year growth 4. Rank of products within category 5. Moving averages ⚙️ Why Use Table Calculations? Show % of total per sub-category, Percent of Total visit www.siradeel.blogspot.com for full details about these sections 🛠️ Basic Table Calculations (No Formula Required) Tableau gives you built-in table calcs through right-click: Try This: Open a new sheet Drag Sub-Category to Rows Drag Sales to Columns Now: Right-click on SUM(Sales) in the view Choose Quick Table Calculation Select: ✅ Percent of Total ✅ Running Total ✅ Rank 🔍 You’ll see how numbers change instantly — that’s a table calc! 📐 Advanced Table Calculations (Using Functions) Let’s create our own calculation now. ✍️ Example 1: Calculate Percent of Total Manually Create a new calculated field: tableau CopyEdit SUM([Sales]) / TOTAL(SUM([Sales])) This divides each row’s sales by the total across the table. Tableau will ask: “Compute using?” ➤ Choose the direction (Table Across, Down, Pane, etc.) 🎯 Understanding Compute Using visit www.siradeel.blogspot.com for full details about these sections 🎓 Tip: Always check "Edit Table Calculation" to set this properly! 🧪 Common Table Calc Functions visit www.siradeel.blogspot.com for full details about these sections 🧪 Hands-on Example: Year-over-Year Growth (YOY) Goal: Compare this year’s sales to last year’s Steps: Drag Order Date to Columns → Choose Year Drag Sales to Rows Create a new calculated field: (SUM([Sales]) - LOOKUP(SUM([Sales]), -1)) / LOOKUP(SUM([Sales]), -1) Format it as percentage 🔍 This shows YOY growth using LOOKUP() to reference the previous year! 🧪 Challenge Task (Try Now!) Goal: Show top 5 sub-categories by running sales total Create a bar chart: Sub-Category vs Sales Sort it in descending order Create a new field: RUNNING_SUM(SUM([Sales])) Add a filter to only show top 5 using INDEX() <= 5 This teaches you to combine ranking + window calculations. 🔄 Table Calculation Best Practices ✅ Use "Edit Table Calculation" to control computation direction ✅ Use tooltips to explain what each number means ✅ Combine with filters for interactive dashboards ✅ Be cautious when changing table structure — table calcs can break

3

1 Comment

🚨 Data Analyst in 2026? Read This. If you only know: SQL Power BI Basic Excel You are NOT safe. Here’s what companies are hiring for right now 👇 ✅ Advanced SQL (CTE, Window Functions, Optimization) ✅ Power BI (DAX, Data Modeling, Performance Tuning) ✅ Snowflake / Cloud Data Warehouse ✅ Python for Data Analysis ✅ Automation using AI tools ✅ Business Storytelling The market has changed. ❌ Companies don’t want “report creators” ✅ They want “decision partners” AI is not replacing Data Analysts. But AI-powered Analysts are replacing traditional ones. In the last 12 months: Hiring slowed Expectations increased Productivity benchmarks doubled So what should you do? 🔹 Learn AI-assisted analytics 🔹 Automate repetitive reporting 🔹 Improve business communication 🔹 Focus on impact, not just dashboards If you’re in: Data Analytics | Business Intelligence | SQL | Power BI | Snowflake Start upgrading NOW. The future belongs to analysts who: ⚡ Think like consultants ⚡ Use AI like assistants ⚡ Deliver measurable business impact Comment “AI” and I’ll share the roadmap I’m following. #DataAnalytics #BusinessIntelligence #SQL #PowerBI #Snowflake #AI #DataAnalyst #CareerGrowth #Analytics #DataScience

7

https://lnkd.in/gY7v8Sjj From Data to Decisions: A Story-Driven Sales Analysis with Tableau Public Today, I’m excited to share a Tableau Public Sales Story I created — a narrative-driven analysis that turns raw sales data into actionable insights. In today’s business world, a chart alone isn’t enough — context and narrative are key to making data meaningful. That’s where Tableau Stories shine: they combine interactive visualizations with captions to guide your audience through your thought process — just like a presentation, but richer and more interactive. Here’s what I focused on: 🔹 Understanding Overall Sales Performance: The opening view provides high-level insight into total sales across categories — a foundation for deeper analysis. 🔹 Analyzing Trends Over Time: By exploring sales and profit trends over multiple periods, we uncover patterns that may relate to seasonality or strategy changes. 🔹 Breaking It Down by Category & Region: Segmenting sales by product categories and regions helps reveal which areas are driving growth and which need attention. 🔹 Highlighting Key Opportunities & Challenges: Through the story’s progression, we surface where profit margins are narrowing despite strong sales and where specific sub-segments may be underperforming. 💡 What makes a story different from a dashboard? It’s not just “what the numbers are” — it’s how they evolve, compare, and connect. Each story point builds on the previous one to enhance understanding. #DataVisualization #Tableau #Analytics #BusinessIntelligence #DataStorytelling

20

3 Comments

🚨 Loan Default Risk Dashboard | Tableau Visualization I'm excited to share my latest data visualization project built in Tableau! 📊 🔍 This interactive dashboard explores loan default risks, offering insights into: ✔️ Comparison of average loan amounts by payment status ✔️ Influence of Annual Income and Interest Rates on loan outcomes ✔️ Loan performance across DTI (Debt-to-Income) ranges ✔️ Repayment behavior by loan term ✔️ Distribution by loan purpose 📌 Key takeaway: Loans categorized as "Not Paid" tend to have higher average amounts and are more prevalent in specific income and DTI bands. Understanding these patterns can help improve credit risk models. 💡 Tools used: Tableau Public Desktop Edition 📁 Data: Lending dataset (public) 🔗 Looking to connect with fellow data enthusiasts and get your feedback! #DataAnalytics #Tableau #LoanRiskAnalysis #DataVisualization #BusinessIntelligence #DataScience #DashboardDesign #LinkedInLearning

19

1 Comment

RMSE isn’t the only metric in Regression. Here’s when to use MAE, R² and MAPE. When I started in Data Science, I thought #RMSE (Root Mean Squared Error) was the metric for regression. But here’s the thing: no single metric tells the full story. 👉 RMSE (Root Mean Squared Error): Punishes large errors more than MAE. Best when big mistakes are very costly (e.g., predicting energy demand too small). 👉 MAE (Mean Absolute Error): Tells you the average error in the same units as your target. Great when you want easy-to-interpret results (“On average, we miss sales by 12 units”). 👉 R² (Coefficient of Determination): Explains how much variance your model captures. Good for understanding model fit but not error magnitude. 👉 MAPE (Mean Absolute Percentage Error): Expresses error as a percentage. Easy to explain to non-technical people (“Our forecast is off by 8% on average”). ⚠️ Watch out: breaks when actual values are close to zero. 💡 Pro tip: Always look at multiple metrics. Each one highlights a different side of model performance. Next time you evaluate a #Regression model, ask yourself: “Do I just want one number, or the full picture?”

82

19 Comments

Tiger Analytics is known for asking some solid, real-world Data Engineering questions during interviews. If you're preparing, here’s a quick snapshot of what they typically expect 👇 💰 Role-wise Compensation (Approx.) Data Engineer: ₹15–20 LPA Senior Consultant – Data Engineering: ₹25–30 LPA 📌 Commonly asked technical questions: 1️⃣ How would you build a highly scalable ingestion pipeline for hourly multi-TB parquet files landing in ADLS using ADF + Databricks? Consider partitioning, pruning, and metadata performance tuning. 2️⃣ Explain the internals of OPTIMIZE and ZORDER in Delta Lake. Also discuss the scenarios where running them may not be beneficial. 3️⃣ Write a SQL query to find customers who purchased in 3 or more back-to-back months. 4️⃣ What’s the performance difference between using Delta Lake CDF and MERGE? When is one a better choice over the other? 5️⃣ How does Spark’s Tungsten engine speed up execution? Talk about off-heap memory and whole-stage code generation. 6️⃣ A Databricks job fails intermittently due to data skew. How do you diagnose it and what steps would you take to mitigate it? 7️⃣ What are the best practices for building resilient ADF pipelines? Think retries, monitoring, alerting, and custom logging. 8️⃣ How would you design a watermark strategy in Spark Structured Streaming when events arrive late or out of order? 9️⃣ Delta Live Tables vs traditional Spark notebooks: When should each be used for production-grade ETL? 🔟 Explain SCD Types 1, 2, and 3. How would you efficiently implement Type 2 using Window functions + MERGE? 1️⃣1️⃣ How do you build enterprise data lineage with Purview integrated across ADF and Databricks? 1️⃣2️⃣ Compare Databricks SQL Warehouse with Azure Synapse Dedicated SQL Pool for analytics workloads. 1️⃣3️⃣ Describe a real scenario where you applied data quality checks (schema, nulls, referential integrity) and how you set up automated alerts. 1️⃣4️⃣ Share an experience where you reduced Databricks costs using autoscaling, spot nodes, or better query optimization. 1️⃣5️⃣ Design a near real-time Kafka → Spark Streaming → Delta pipeline. How do you guarantee exactly-once processing and handle replay logic? 👉 If you found this helpful, hit LIKE 👍 and RESHARE 🔁 so more people can benefit from it! 📍 𝗙𝗼𝗹𝗹𝗼𝘄 𝗺𝗲 𝗵𝗲𝗿𝗲 𝗳𝗼𝗿 𝗱𝗮𝗶𝗹𝘆 𝗦𝗤𝗟, 𝗗𝗮𝘁𝗮 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴, 𝗮𝗻𝗱 𝗝𝗼𝗯 𝘂𝗽𝗱𝗮𝘁𝗲𝘀: 👉 https://lnkd.in/dhYpfrS7 𝗜 𝗵𝗮𝘃𝗲 𝗽𝗿𝗲𝗽𝗮𝗿𝗲𝗱 𝗖𝗼𝗺𝗽𝗹𝗲𝘁𝗲 𝗜𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄 𝗣𝗿𝗲𝗽𝗮𝗿𝗮𝘁𝗶𝗼𝗻 𝗚𝘂𝗶𝗱𝗲 𝗳𝗼𝗿 𝗗𝗮𝘁𝗮 𝗘𝗻𝗴𝗶𝗻𝗻𝗲𝗿𝘀/𝗗𝗮𝘁𝗮 𝗔𝗻𝗮𝗹𝘆𝘀𝘁𝘀. 👉𝗚𝗿𝗮𝗯 𝗶𝘁 𝗵𝗲𝗿𝗲(𝟭𝟬𝟬% 𝗥𝗲𝗳𝘂𝗻𝗱): https://lnkd.in/giHUMNx9

110

4 Comments

A quick analysis in Tableau ended up surfacing an insight I didn’t see coming… I was analyzing our sales and operations dataset, and one comparison immediately stood out. Even though Sales brought in ₹21.18L in revenue, Operations generated ₹13.82L with almost the same employee count (28 vs 20) but their profit contribution was noticeably lower, especially when compared to branches like Goa, which delivered ₹10.48L in profit despite having a smaller expense share. Using Tableau’s bar charts and pie visuals, I mapped revenue, expenses, profit, and satisfaction side by side. The interesting pattern? Finance contributed ₹16.39L in revenue with one of the lowest employee counts (19), yet maintained the highest satisfaction score (1,550), hinting at strong process efficiency. It was a good reminder that headcount is not equal to performance and revenue is not equal to profitability. Curious if others have seen similar cross-department efficiency gaps in their Tableau dashboards.

3

New Tableau Project: Stock Market Analysis Dashboard Just wrapped up an exciting data visualization project — Stock Market Analysis — built entirely in Tableau! Using real stock data from Excel, I explored how Open, Close, EMA (Exponential Moving Average), and RSI (Relative Strength Index) influence market behavior over time. Highlights of the Dashboard: * Trend analysis of stock prices (Open vs. Close) * RSI-based overbought/oversold insights * EMA vs Close comparison for momentum tracking * Yearly and monthly performance breakdown * Summary stats like Median, Max, Min & Volatility Key Insight: Stocks showing RSI > 70 often signaled short-term overbought conditions — a strong hint for upcoming corrections or consolidations. GitHub : https://lnkd.in/gPrBRX_q

13