Let’s talk

Case Studies

I break down the problems I worked on, the design decisions I made, and what I learned from the results.

Building a Domain-Specific Medical Text Summarization System

Medical question-answer content is dense, technical, and time-consuming to interpret for both clinicians and patients.

Primary MetricAchieved 88%+ summarization accuracy.
Know full study

Designing a Multimodal AI System for Mental Health Screening

Early detection of mental health indicators requires combining emotional facial cues with textual sentiment signals.

Primary MetricImproved prediction robustness through ensemble modeling.
Know full study

Implementing RAG in a Production-Ready AI Complaint Management System

Traditional complaint systems lack intelligent prioritization and contextual response generation.

Primary MetricReduced response latency through retrieval optimization.
Know full study

From CNN to YOLO: Improving Crop Disease Detection with Object Detection

Baseline CNN model struggled to capture contextual image features, limiting accuracy.

Primary MetricAchieved 83%+ disease classification accuracy.
Know full study