Profile banner
Ravi Teja Maddhini

Ravi Teja Maddhini

Data scientist focused on computer vision and NLP

ravi-maddhini
Washington, United States
Bangalore, India
Joined March 2026
Favicon imagegithub.com
Favicon imagelinkedin.com
Favicon imagequora.com
Favicon imagefacebook.com
Favicon imageinstagram.com

Network

3.5K connections
🎓
Georgetown MBA Students
🎯
Product Management Strategy
🤖
AI ML Data Leaders
🏛️
Georgetown Faculty Staff
💰
Emburse Professionals
💻
Samasra Soft Engineers
☁️
...

Summary

Practical NLP and computer-vision practitioner: repositories and profile describe work on sentence embeddings (SBERT), word co-occurrence matrices, sentiment-analysis hackathon projects, and computer-vision classification exercises, indicating hands-on experience building NLP and CV models. github+3
Industry experience across data and ML roles: public profiles list roles at multiple companies (Emburse, Phenom People, OpsMx, Samasra Soft Inc), showing a career trajectory through data-science and machine-learning positions. theorg+1
Early-stage technical leadership: listed as a founding member and technical lead at Samasra Soft Inc (2021–2024), indicating experience in technical leadership and startup environments. weekday
Open-source and community contributor: maintains multiple public GitHub projects and has answered technical questions on Data Science Stack Exchange, demonstrating engagement with the technical community and public-facing contributions. github+1

Work

Education

Projects

Writing

Is it possible to predict sentiment of unlabelled dataset using BERT?

October 1, 2020

Answer recommending annotating a subset to fine-tune BERT or using open-source sentiment datasets and transferring a classifier to unlabeled data.

Favicon imagedatascience.stackexchange.com