Nirmal Budhathoki's Linkedin Analytics

🚀 From a Passion to a Billboard Moment! Proud to share that I’ve been featured once again by topmate.io on the New York city's times-square billboard as one of their elite mentors. While I’ve always believed in giving without expectations, moments like these fuel my purpose even more 🙏 A quick rewind: 🔹 Started my Topmate journey in late '23, became fully active in early '24 🔹 Completed 1500+ bookings, with ~700 being my free DS tips document 🔹 Still 800+ real mentorship sessions with people around the world 🔹 Maintained a 4.9/5 ratings overall 🔹 And most importantly — used the funds to support women’s education in Nepal through the DevMiri Women Education Program ❤️ To everyone who booked a session, downloaded a product, or simply rooted for me- this recognition is yours too. For me, YOU are the real elites. This is just the beginning. Many more milestones to come. 💫 Let’s keep learning, mentoring, and making an impact together. #datascience #mentoring

Excited to join University of Washington today to speak with MSIM students about navigating the data science career path. I know the current job market feels tough. The graduating students will have more nervousness than excitement. I graduated during the 2009/10 recession- so I truly understand that mix of anxiety and uncertainty. From the military to Microsoft, my journey had its share of challenges. While we can’t control everything, we can plan and adapt. I’ll be sharing my story, practical tips, and answering all your questions. See you all there today! Thank you Association of Information Management Students at University of Washington (UW AIMS) for organizing this and having me. #datascience #conference

Today is not just a long weekend. It’s a day of remembrance. As a former service member, I’ve seen how Memorial Day is often misunderstood. It’s not about honoring those in uniform who are still with us, it’s about remembering those who never made it back. We celebrate Veterans day in November to honor all service members- living or deceased. Today is about the lives lost in service to freedom. And the families who have sacrificed. I carry the memory of friends who gave everything. And I also know some of those who made it home but faced a different kind of war inside their minds. Mental health is a battlefield too. I’ve lost friends not to combat, but to the silent wounds they couldn’t heal. So today, as we get a day off or spend time with loved ones, let’s not forget why this day exists. Honor. Remember. And take care of yourself and those around you. #randomtopics #memorialday

🚨 Planning to kickstart your Data Science career in 2025? I recently gave a talk covering some crucial (and often overlooked) tips for anyone entering the field- especially new grads. Covered in the presentation: 🔹 3 types of Data Science roles (and how they differ) 🔹 Common myths in DS you must stop believing 🔹 How to tailor your resume beyond keywords 🔹 Why networking is your secret job-search weapon 🔹 The real strength of a solid project portfolio I’ve shared the PDF of the presentation. Feel free to grab it, and please share it with your network. #datascience #machinelearning

Meta Data Science Interview Prep Lately, I have been talking to my mentees to help them with Interview Prep calls. As a data-driven person, I find a positive sign in job market when I get more bookings for interview preps as compared to other services :) Last time I posted some sample questions from DoorDash, and I received lots of positive messages- saying it was very helpful. As I talk to many folks about interview preps, I gather some great insights. I've decided to share it to help others who are now preparing or in future. DS role at Meta is mostly product focused- heavy on SQL and product analytics. Those DS roles are not ML focused. If you want to align with ML, I think the ML engineer roles are better fit. The first technical round for Meta's DS role- is about SQL plus case study. Usually 2-3 SQL built on each other- mostly starting easy to medium and end with advanced one. Below are some samples- again it is based on my conversation with folks who completed their interviews. Given a schema of table: user_sessions user_id (int): Unique user ID session_id (string): Unique session ID session_start (datetime): Timestamp session started session_end (datetime): Timestamp session ended device_type (string): 'iOS', 'Android', 'Web' country (string): Country of the user event_type (string): 'view', 'like', 'comment', 'share' event_time (datetime): Timestamp of each event Below are SQL questions: 1. Find the top 3 countries by average session duration over the last 7 days. 2. List the top 5 device types by engaged user count. An "engaged user" is the ones who had at least 2 sessions and performed a 'share' event at least once in the last 14 days. Case Study: Meta recently launched a new comment filter feature that auto-hides spam comments. After rollout, there's a 2% drop in user comments overall. Leadership is concerned it might be harming engagement. How do you analyze this ? I want you to give a try, practice and learn. Please share it with your network. I post data science insights daily. Follow for more. #datascience #meta #interview

10,000 YouTube Subscribers! Not a biggest number, but definitely a meaningful one. I believe in small wins :) Every video I’ve shared took hours of thought, editing, and most importantly-intention to help someone learn, grow, or get inspired in their data science journey. This milestone isn’t just a number. It’s a reminder that consistency compounds, and value finds its audience- may be slowly, but surely. To those who’ve watched, liked, commented, or shared- Big thank you! If you haven’t checked it out yet, and you’re into data, ML, or career stories with a real-world touch—come say hi: [https://lnkd.in/gBP2C_NY] Let’s keep learning, one byte at a time. #datascience #machinelearning #learning

🚀 Data Science Interview Coming Up? Keep This by Your Desk 🐼 One of the most googled search are pandas commands, no doubt. It is not about memorizing the syntax, but it is about using it when needed at finger tips. I prepared a quick cheatsheet what I would print and stick next to my desk. Whether its for your interviews or daily use, it will be handy. This cheatsheet covers: ✔️ Filtering, Grouping, Merging ✔️ Handling Missing Data ✔️ String and DateTime operations ✔️ Must-know patterns like ranking, case logic, and more! Whether you're preparing for interviews or working with real-world data, this is something you’ll want to keep within your arm's reach. Grab your free PDF copy as attached. I ask you only one favor. Please help me spread the words by sharing it to your network. Together we learn and grow. #datascience #pandas

A good Data Scientist is a good debugger too ... And no, I’m not talking about just debugging code. Real-world data science isn't just about modeling. It's about debugging everything: 🔹 Debugging upstream data issues 🔹 Debugging broken assumptions 🔹 Debugging pipelines that worked yesterday 🔹 Debugging why your prod model went from F1-score 0.85 to 🤷 This is where logging becomes your best friend. As one of my recent podcast guests said: "If you're not logging your model's behavior in production, you're just hoping it works." Some of the popular Python logging tools to look at: 🔹 logging (standard, but needs setup) 🔹 loguru (elegant, simple, powerful- go check it out) 🔹 structlog (for structured logs in microservices) 🔹 sentry / opentelemetry for advanced monitoring In production, logs are your only eyes into what’s happening. They tell the story data can’t- until it’s too late. So, next time you build that model, don’t just print accuracy. Log the journey. Debug the process. Because a good data scientist isn’t just a model builder, they’re detective in disguise. 🕵️‍♂️ Happy Friday, detectives! #datascience #machinelearning

The House of Data Science- which door would you open? In one of my YT videos, I break down the three common roles under the Data Science umbrella: 🔹 BI or Analytics-Focused DS: Think dashboards, KPIs, and business health checks. Skills: SQL, data visualization, data engineering basics, storytelling with data, business acumen. 🔹 Product-Focused DS: The decision helper for product teams- should we launch this feature? Is this experiment a win? Skills: SQL, A/B testing, experimentation design, behavioral metrics, KPI optimization. 🔹 ML-Focused DS: Model builders and deployers. Turning data into predictions, automations, and intelligent systems. Skills: Python (more), SQL (some), ML/DL frameworks, PySpark, MLOps, cloud tools. Check out the full breakdown in the video: [https://lnkd.in/gS5qrGHH] #datascience #roles

I’ve done countless talks and presentations on data science and career growth. But this one was truly special. This time, I wasn’t presenting to professionals- I was presenting with my daughter, to her classmates, about Nepal. No slides. Just placards, fun facts, and curious little minds. We shared 5 things about Nepal 1. Home to Mount Everest – the highest peak in the world 2. Birthplace of Lord Buddha – Lumbini 3. The only non-rectangular national flag 🇳🇵 4. Living goddess tradition – Kumari 5. And of course—our rich culture and warm hospitality It was simple. It was joyful. And honestly, it was one of the best experiences I’ve had. And ya I had to put Nepali hat- Dhaka Topi 😊 Sometimes, the most meaningful presentations happen far away from conference rooms. #randomtopics #nepal #prouddad

For data science roles- the level of coding depends on companies. Based on my own experience, for the most part- coding rounds include: 1. SQL: if you are doing analytics type data science roles, you will be tested for SQL as a part of coding round. Example: Data Scientists who are directly attached to product managers. 2. Python (DSA): if you are doing core machine-learning type data science roles, then your will be tested for data structure and algorithm (DSA). This is where we have to learn and prepare for some DSA. This post will cover the number 2 from above. In my own experience, and having conversation with many other mentees, as well as my peers, DSA rounds for ML data scientists are not as heavy as software engineers. Therefore, I suggest covering the easy-medium level Leetcode problems to practice. However iff you are interviewing for Machine Learning Engineer (MLE) role, then you have to go more depth. MLE is actually SWE role working in ML systems. For Data Scientists side- I prepared a playlist covering 30 coding questions, that will help you prepare for data science coding round. Link on comments. Tip: Mostly I always hear leetcode as single source of prep, but I have found needcode as very useful. Navdeep Singh has done an amazing job 👏 #datascience #machinelearning