Abhinav Ravi - AI Researcher

About

I'm Director of AI research at Krutrim AI Labs, working on cutting-edge multilingual language models designed for India's diverse linguistic landscape. My research focuses on making AI accessible to over a billion people through culturally-aware and linguistically-inclusive systems.

Originally from Muzaffarpur, Bihar, I completed my schooling at DAV Public School and DPS Bokaro Steel City before pursuing Computer Science at IIT Guwahati. There, I developed a passion for machine learning, computer vision, and natural language processing. My work spans across conversational AI, multi-modal systems, and deep metric learning.

Beyond research, I'm deeply curious about the frontiers of human knowledge and enjoy exploring creative pursuits through photography and visual storytelling.

Beyond Research

When I'm not training models or writing papers, I enjoy capturing moments and exploring creative pursuits. Here's a glimpse into my world beyond AI research.

Education

Bachelor of Technology (B.Tech)

Computer Science & Engineering, CPI: 8.73/10

Indian Institute of Technology (IIT) Guwahati

Graduated with honors in 2016. Focused on Machine Learning, Computer Vision, and Algorithms. Completed multiple research projects under faculty guidance.

2012 - 2016

Senior Secondary Education (Class 11-12)

CBSE Board, 92%

Delhi Public School (DPS), Bokaro Steel City

Specialized in Science stream with Mathematics, Physics, and Chemistry. Developed strong analytical and problem-solving foundations.

2010 - 2012

Secondary Education (Class 1-10)

CBSE Board, 96%

DAV Public School, Muzaffarpur, Bihar

Completed foundational education with consistent academic excellence. Participated in various science and mathematics competitions.

2002 - 2010

Research & Publications

Krutrim LLM: Multilingual Foundational Model for over a Billion People

A. Kallappa, P. Kamble, A. Ravi, et al.

Introducing a 2 trillion token multilingual model designed for India's linguistic landscape with the largest known Indic dataset. Matches or exceeds LLAMA-2 performance while ensuring balanced representation across dialects.

arXiv 2025

Chitrarth: Inclusive Vision-Language Model

A. Ravi and team

A state-of-the-art multimodal VLM targeting 10 prominent Indian languages, integrating visual reasoning with multilingual capabilities. Introduces BharatBench for comprehensive evaluation.

Krutrim AI Labs

Detection and Localization of Metastasis Cancer Using Pathology Images

A. Ravi and collaborators

Computer vision system for automated cancer detection. Won 2nd prize at PAN-IIT AI Conclave at IIT Delhi.

PAN-IIT AI Conclave

Color Variants Identification in Fashion E-Commerce

A. Ravi et al.

Deep visual representation learning framework for identifying fashion products that match in design but differ in color, improving customer experience in e-commerce.

Research Publication

Experience

Director, AI Research

Krutrim AI Labs, Bangalore

2023 - Present

Leading research on multilingual LLMs and vision-language models for Indic languages. Core contributor to Krutrim LLM pre-training, fine-tuning, and alignment. Building foundational AI systems for over a billion people.

Staff Data Scientist

Myntra (Flipkart Group)

2017 - 2023

Worked on computer vision and deep learning systems for fashion e-commerce. Developed recommendation engines, visual search systems, and product matching algorithms. Awarded "High Achiever" for exceptional DS model performance.

Software Developer Intern

CRM Solutions

May 2015 - July 2015

Built intelligent NLP-based auto-classifier and auto-replier system for customer relations. Tackled complex challenges including context maintenance and intent mining using supervised machine learning.

Research Intern - Computer Vision

IIT Guwahati

May 2014 - July 2014

Under Dr. P.K. Das. Developed smart lighting system using OpenCV and computer vision to automatically control lighting based on human presence detection.

Research Interests & Skills

Machine Learning

Deep Learning NLP LLMs Transformers

Computer Vision

Image Recognition VLMs OpenCV Metric Learning

AI Research

Multilingual AI Conversational AI Multi-Modal AI RAG

Engineering

Python PyTorch TensorFlow CUDA

Get In Touch

📧 ravi.abhinav3@gmail.com

📍 Bangalore, India | From Muzaffarpur, Bihar

🎓 Google Scholar