Graduate student researcher · Software engineer

Prajwal Kadri

Profile

Working in AI research and software development while exploring other fields.

I am a graduate student pursuing an M.S. in Intelligent Information Systems at Carnegie Mellon University’s Language Technologies Institute. Previously, I worked on full-stack client-platform systems at Visa. My research spans multilingual speech, computer vision, and machine-learning applications, with a current focus on language diarization.

Selected experience

01 / 04
2026 — PresentCarnegie Mellon University

Graduate Student Researcher

Conducting research in the Multimodal Signal Processing Lab with Prof. Carlos Busso on language diarization, developing and evaluating methods for multichannel, multi-speaker settings.

Speech TechnologyMachine LearningLanguage DiarizationMultimodal ML
2023 — 2026Visa

Software Engineer

Implemented workflow-driven form modules and Spring Boot microservices while remediating security vulnerabilities in containerized applications. Also implemented data masking and database encryption within six weeks.

JavaSpring BootAngularTypeScriptMicroservicesDockerCI/CD
2022Visa

Summer Intern

Developed a full-stack configuration editor that enabled business teams to update form fields and tooltips without developer intervention, reducing UI-change turnaround time from 70 days to seconds through data-driven, real-time configuration and publishing workflows.

JavaAngularPostgreSQLFull-stack Development

Research & publications

Google Scholar ↗
01

Language Detection in Overlapping Multilingual Speech: A Focus on Indian Languages

Developed a separation-first language-identification pipeline for overlapping Indian-language speech, using a synthetic dataset derived from VoxLingua107.

↗ 02

Cardiovascular Disease Prediction Using Machine Learning

Evaluated five machine-learning classifiers for early heart-disease prediction from physiological and demographic features.

↗ 03

Multi-Vehicle Tracking and Speed Estimation Model Using Deep Learning

Combined YOLOv5 detection, multi-object tracking, and calibrated pixel-to-distance conversion for real-time vehicle speed estimation.

↗ 04

Device to Device Pairs Sharding Based on Distance

arXiv preprint on distance-based sharding for device-to-device pairs.

↗

Selected projects

GitHub ↗
01

D2D Pairs Sharding Based on Distance

Implemented distance-based D2D clustering and SINR calculation using an ns-3 simulation environment.

Networks · ns-3 · 5G ↗
 02

CAPTCHA Solver

Built CNN-based CAPTCHA solvers and a web application for interactive prediction demos.

Machine Learning · CNN ↗
 03

NITK Social Media App

Created a campus platform for student discussions, anonymous questions, and organization event discovery.

Full-stack · Angular · Node.js ↗

Education & recognition

04 / 04

Graduate study

Carnegie Mellon University

Master of Science in Intelligent Information Systems
Language Technologies Institute · Expected Dec 2027

Undergraduate study

National Institute of Technology Karnataka

Bachelor of Technology, Information Technology