Fine-Tuning LLaMA2 for News-Aware Conversational AI

I fine-tuned LLaMA2 for news classification without enterprise-grade hardware. The interesting part? LLaMA2 can be trained to be a news expert without losing its friendly conversational skills.

Check out the full article here on Substack.

Understanding Speech Recognition: A Hands-On Guide to Phoneme Classification

Every voice assistant we’ve used—like Siri or Alexa—works by breaking speech into its basic sound units, called phonemes. In this article, I share my learnings in tackling phoneme classification.

Check out the full article here on Substack.

Transform Your Data into Interactive Stories with Streamlit

Imagine turning complex datasets into clear, compelling narratives that speak directly to decision-makers. In data science, this is an essential practice. Streamlit stands out as a useful tool that transforms traditional, static data notebooks and reports into interactive, engaging experiences.

Check out the full article here on Substack.

Teaching Machines to Read: A Case Study in AI-Powered Banking

I recently worked on a case study exploring how AI could revolutionize check processing at a major bank. The scenario involved a financial institution processing 100,000 checks daily, where we designed a machine learning system to automatically read and process handwritten checks – transforming a task that traditionally required manual entry and cost millions in labor.

Through detailed analysis, the proposed AI solution demonstrated the potential to slash processing time by 90% and save $1.8M annually. While this was a theoretical exercise, it perfectly illustrates how artificial intelligence can modernize traditional banking operations, freeing up human staff for more meaningful work while improving accuracy and customer satisfaction.

Check out the case study at my GitHub repository here.

Exploring Humor Detection in Text with NLP

I’m excited to share an NLP Primer focused on humor detection using the ColBERT dataset from Kaggle. In this hands-on project, I dive into using Natural Language Processing (NLP) techniques to identify humorous content in text, a fascinating area with applications ranging from content moderation to chatbots and more.

Read the full article here on Substack.

Introducing My AI/ML Primer on Classification

Welcome to my first AI/ML primer, where I dive into classification techniques through a hands-on project focused on Client Subscription Prediction. In this project, I've developed a machine learning model to predict which clients are most likely to subscribe to a term deposit based on a range of features, including historical data from marketing campaigns.

Read the full article here on Substack.