Welcome to my digital playground! I'm Veda Kailasam and I'm totally geeking out over Artificial Intelligence and Machine Learning.
Picture this: me, a wide-eyed grad student at the University of Illinois Urbana-Champaign, diving deep into the world of Machine Learning, getting my hands dirty with all sorts of cool projects. I'm talking late-night coding sessions fueled by nothing but a hefty dose of caffeine and a relentless curiosity to unravel the mysteries of AI. I'm all about turning cutting-edge tech into real-life solutions.
My journey has been a mix of serious study and practical experience. I've been working closely with Professor David Dalpiaz, the brilliant mind who took me under his wing as a Graduate Teaching Assistant, helping to shape the minds of future data scientists. I've been exploring ways to use fancy AI models like LLMs to personalize tools in data science. How cool is that?
Before my grad school gig, I was over at VerSe Innovation, where I helped shake things up with Dailyhunt's content recommendation systems. We used some pretty advanced techniques like two-tower models, LDA, and Locality-Sensitive Hashing, and the results were pretty sweet. My tech stack reads like a toolkit for the modern AI practitioner and it is packed with Python, Pyspark, LLMs, SQL, GCP, AWS, Git, Docker, and more. It was a wild ride, but seeing the impact of my work on user engagement and content relevance was worth every sleepless night.
But hey, life isn't all about algorithms and models. When I'm not knee-deep in code or buried in research papers, I've got a bunch of other hobbies to keep me busy. You'll often find me lost in the pages of a good book. And when the mood strikes, I'll trade my books for the dance floor, where I can let loose to the rhythms of Bollywood music.
So, come join me on this wild ride of innovation and discovery! Every line of code we write is another step toward unlocking the potential of AI to change the world. Check out my recent work below. Let's do this!
A multi-stage re-ranking approach for LLMs, enhancing listwise passage ranking by refining initial data order and compressing text to address context window limitations, thereby improving retrieval efficiency and accuracy.
An AI-powered video search and extraction tool, enhancing video content discovery with increased search accuracy and reduced extraction times.
Enhancing photo gallery organization with advanced face recognition and clustering technologies to achieve high accuracy and robust model generalization.
Stay tuned for more on my projects, work, and other exciting endeavors—coming soon!
Working with Professor David Dalpiaz on CS 498: End-to-End Data Science and CS 307: Model and Learning in Data Science tailoring educational content to improve learning outcomes.
Revolutionized the recommendation engine with advanced PySpark and NLP techniques by crafting dynamic end-to-end content processing pipelines.
Automated news article grouping and enhanced user engagement through innovative methods like knowledge graphs and community detection, delivering personalized news articles to millions of users through the Dailyhunt platform.
Focused on face recognition and text detection, using Transformer-based models to develop accurate face attribute classifiers and improve OCR capabilities. Enhanced image-based text processing and facial analysis, and contributed to the design and implementation of AI infrastructure and data pipelines, boosting system reliability.
Relevant Coursework: Using LLMs aka ChatGPT, Computer Vision, Applied Machine Learning, Text Information Systems, Data Science Algorithms
Relevant Coursework: Algorithms, Data Structures, Machine Learning, MVC through Python
