Adithya Kolavi
Research Fellow at Microsoft
Bengaluru, Karnataka, India
Summary
🤩
Rockstar🎓
Top SchoolAdithya Kolavi is a research engineer with 9 years of hands-on experience building production-ready AI systems, currently focused on vision-language models and agentic systems at Hugging Face. He has led end-to-end projects from novel research to deployed products—reducing time-to-production for enterprise LLM deployments and fine-tuning over 50 models for diverse domain and language needs. At CognitiveLab he helped build India’s first Kannada bilingual LLM and an open-source data ingestion platform (OmniParse), reflecting a strong commitment to accessible AI. His background spans internships and research roles at Microsoft, Apple, and IISc, combining practical cloud deployments with deep model work like Vision Transformers and RAG optimization. An active open-source contributor and educator, he publishes tooling and tutorials to lower the barrier to applied AI and has experience shipping backend SDKs for document-to-GenAI workflows. Colleagues describe him as the kind of engineer who turns “wild ideas” into reliable systems while keeping deployment costs and evaluation rigor front and center.
9 years of coding experience
3 years of employment as a software developer
Bachelor of Technology - BTech, Computer Science, Bachelor of Technology - BTech, Computer Science at PES UniversityNights and Weekend S4, Nights and Weekend S4 at buildspace
Github Skills (17)
fastapi10
python10
apidoc10
back-end-development10
machine-learning10
svm10
mask-rcnn10
deep-learning10
api10
tensorflow10
http-request10
neural-network10
faster-rcnn10
backend10
base649
Programming languages (11)
Github contributions (5)
adithya-s-k/omniparse
Jun 2024 - Jan 2025
Ingest, parse, and optimize any data format ➡️ from documents to multimedia ➡️ for enhanced compatibility with GenAI frameworks
Role in this project:
Back-end Developer
Back-end DeveloperContributions:6 PRs, 55 pushes, 42 comments in 7 months
Contributions summary:Adithya primarily contributed to the development of the OmniParse Python SDK client. Their work involved the initial implementation of the client, including creating an `OmniParse` class and methods for loading and converting PDF files to markdown and images. They also added a base server with router functionality and a model loader, indicating work on the project's core parsing and processing backend. The commits demonstrate a focus on setting up the API for document processing and image extraction and conversion.
ingestion-apiocromniparserparse-serverparser-library
abhisheks008/DL-Simplified
Aug 2022 - Sep 2022
Deep Learning Simplified is an Open-source repository, containing beginner to advance level deep learning projects for the contributors, who are willing to start their journey in Deep Learning. Devfolio URL, https://devfolio.co/projects/deep-learning-simplified-f013
Role in this project:ML Engineer
ML EngineerContributions:51 commits, 3 PRs, 11 comments in 1 month
Contributions summary:The user, Adithya S K, implemented and tested an SVM model for shoe classification, creating an `SVM_shoe_classification.ipynb` notebook. They further developed a CNN shoe classifier documented in `CNN_shoe_classification.ipynb` and added an output image to `SVM_shoe_classification.ipynb`. This suggests an involvement in model building, evaluation, and potentially experimentation with different machine learning architectures (SVM and CNN) within the domain of image recognition.
pytorchannpythonssoc2022deep-learning
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.