Hey, I’m Vishnu Thakral.
A Shutterbug
I turn data into insights, ideas into stories, and moments into memories. In short, I'm a blend of technology, creativity, and curiosity with a pinch of kindness - all wrapped into one dynamic journey.
When I'm not building ML models or analyzing data, you'll find me behind a camera capturing life's beautiful moments, or writing about machine learning to help others on their journey. I believe in making complex concepts accessible and sharing knowledge that empowers the community.
My journey in machine learning and data science started with curiosity - asking "what if?" and "how can we make this better?" This curiosity led me from building my first neural network for object detection to leading ML teams and teaching others.
I'm passionate about democratizing AI knowledge. That's why I created MLGyaan.com - a platform where I write accessible articles about machine learning, breaking down complex concepts into digestible insights for the community. Whether it's explaining the difference between methods and functions, or diving deep into statistical concepts, I believe learning should be approachable and engaging.
Beyond code and algorithms, I'm a storyteller at heart. I capture moments through photography, turning everyday scenes into visual narratives. This creative side complements my technical work, helping me see problems from different angles and communicate complex ideas more effectively.
My mission? To bridge the gap between cutting-edge AI research and practical, real-world applications - all while making the journey enjoyable and accessible for everyone.
NYU Self Drive
Built a vision-based AI model for AWS DeepRacer, enhancing autonomous racing with reinforcement learning.
SecuroDrive (Driver Drowsiness Detector)
Developed a real-time AWS-based drowsiness detection system.
Lip Reading Assistant
Deep learning-based lip-reading model for speech-impaired individuals.
Stay Recommender
Enhanced Airbnb experience with safety filters, restaurant insights, and AI-powered rating predictions.
Fairprep
Open source modeling tool for ensuring fairness in Machine Learning
Smart Office Access Optimization
Developed an AI-powered application to ensure safe and efficient office space utilization as employees returned to work post-COVID.
Vansh, V., Chandrasekhar, K., Anil, C.R., Sahu, S.S. (2020).
Improved Face Detection Using YCbCr and Adaboost.
In: Behera, H., Nayak, J., Naik, B., Pelusi, D. (eds) Computational Intelligence in Data Mining.
Advances in Intelligent Systems and Computing, vol 990. Springer, Singapore.
Paper Link
I'm passionate about making machine learning accessible to everyone. That's why I created MLGyaan.com - a platform where I write articles to help the ML community learn and grow. I also share insights on Medium. Here's a selection of my recent work:
Prompt Engineering Demystified: From Zero to Hero
You know that feeling when you ask ChatGPT something and get back a response that's... technically correct but not quite what you wanted? Learn the art and science of prompt engineering to get the results you need.
Metrics that Matter: Beyond Accuracy
From accuracy traps to fairness and security, a definitive guide to ML metrics with real-world cases and equations. Learn when to use which metric and why accuracy alone isn't enough.
A/B Testing Demystified - Part 2: Beyond the Basics
Explore advanced A/B testing - power, sample size, Bayesian methods, SRM checks, and real-world pitfalls. Take your experimentation skills to the next level.
A/B Testing Demystified...Part 1: The Science of Fair Comparisons
Behind every A/B test is a courtroom trial of statistics. Let's explore how hypotheses, errors, and sample size shape the truth. Master the fundamentals of experimentation.
Mean, Median, Mode: The Three Musketeers of Misleading Stats
Everyone lies with statistics. Here's how to catch them in the act - a fun guide to understanding when averages mislead and how to choose the right measure of central tendency.
The Art of Asking Good Questions
Asking the right questions is an essential skill in data science. Learn how to frame questions that lead to actionable insights and better decision-making.
Sips and Stats: Seattle's Best Coffee Spots Revealed
We used real-time data, machine learning, and a touch of caffeine science to explore and reveal the best coffee spots in Seattle using Yelp data and clustering algorithms.
Deciphering Method Calls vs. Function Calls
Methods are bound to class instances and operate on object attributes (e.g., a.lower() for strings). Functions are standalone, reusable blocks operating on input arguments (e.g., calculate_rectangle_area(length, width)). Since lower() is a string method, calling lower(a) results in an error; use a.lower() instead.
Unlocking Insights with Descriptive Statistics
Descriptive statistics simplify complex data using key measures like mean, median, and standard deviation. They reveal patterns, trends, and variability, making data-driven insights clearer and more actionable.
Through my lens, I capture moments that tell stories - from urban landscapes to nature's beauty. Photography is my way of seeing the world differently and sharing those perspectives.
Machine Learning & AI
Data Science
Cloud & DevOps
Web Technologies
New York University
M.S. in Computer Science (2019 - 2021)
Birla Institute of Technology, Mesra
B.Tech in Electronics & Communication (2014 - 2018)
Data Science Bootcamp - New York University
2019 - 2021
Led comprehensive hands-on bootcamps covering Python programming, machine learning fundamentals, and data science workflows. Designed curriculum and practical exercises that helped students build real-world projects. Focused on making complex concepts accessible through interactive learning and real-world examples.
Impact: Trained 2000+ students in data science fundamentals, with many going on to successful careers in tech.
Amazon ML University
2024 - Present
Instructor for advanced courses in Natural Language Processing (NLP) and Responsible AI. Developed course materials covering transformer architectures, BERT, GPT models, and ethical AI practices. Conducted workshops on bias detection, fairness metrics, and model interpretability.
Impact: Trained over 150 Amazon employees, enabling them to build more ethical and effective ML solutions.
MLGyaan.com - Community Education
2024 - Present
Created and maintain MLGyaan.com, a platform dedicated to making machine learning accessible. Write articles, create tutorials, and share knowledge with the global ML community. Focus on breaking down complex topics into digestible content that helps both beginners and experienced practitioners.
Impact: Reaching thousands of learners worldwide, helping democratize AI education.
Feel free to reach out for collaborations, opportunities, or just a chat.