Hi, I am Hrithik Sharma, an M.Tech student at IIT Delhi specializing in Artificial Intelligence and Machine Learning.
My work focuses on building practical, research-oriented, and reproducible solutions across:
- Machine Learning and Deep Learning
- Natural Language Processing and Transformers
- Cloud Computing with Azure, Docker, and FastAPI
- Recommendation Systems
- Graph Machine Learning
- Data Structures and Algorithms
I enjoy developing systems that combine strong engineering fundamentals with real-world usability.
Contributing production fixes and regression coverage to Kornia, an open-source differentiable computer vision library built on PyTorch.
| Status | Contribution | Pull Request |
|---|---|---|
| ✅ Merged | Implemented Brown–Conrady, Kannala–Brandt K3, and Orthographic camera models, including distortion/projection plumbing, intrinsic matrices, batched project/unproject support, tests, docs, and export-support coverage. | kornia/kornia#4377 |
| ✅ Merged | Corrected the analytic Kannala–Brandt distortion Jacobian, including the small-radius branch and reduced-precision numerical stability. | kornia/kornia#4368 |
| ✅ Merged | Fixed rad2deg / deg2rad for integer tensors and preserved float64 precision while retaining ONNX-compatible tensor arithmetic. |
kornia/kornia#4358 |
These contributions involved reproducing upstream defects, implementing focused fixes, adding regression tests, validating across multiple dtypes/backends, responding to maintainer review, and getting the changes merged into main.
|
AI-powered study assistant supporting document upload, quiz generation, flashcards, roadmaps, summaries, doubt solving, and local AI workflows. Technology: Python, Streamlit, Ollama, PDF tooling |
Safety-first computer vision framework for traffic violation evidence generation with structured outputs and manual review fallback. Technology: Python, Computer Vision, OCR, YOLO |
|
Low-resource neural machine translation pipeline for Pashto-to-English and Pashto-to-Hindi translation. Technology: PyTorch, Hugging Face, Transformers, NLP |
Structured problem-solving repository with topic-wise solutions, explanations, dry runs, and complexity-focused implementations. Technology: Python, Algorithms, Data Structures |
| Artificial Intelligence / Machine Learning | 🟨🟨🟨🟨🟨🟨🟨🟨🟨⬜ | 90% |
| Cloud Computing | 🟦🟦🟦🟦🟦🟦🟦🟦⬜⬜ | 80% |
| Deep Learning | 🟧🟧🟧🟧🟧🟧🟧🟧⬜⬜ | 80% |
| NLP and Transformers | 🟩🟩🟩🟩🟩🟩🟩🟩⬜⬜ | 80% |
| Recommendation Systems | 🟪🟪🟪🟪🟪🟪🟪⬜⬜⬜ | 70% |
| Graph Machine Learning | 🟫🟫🟫🟫🟫🟫🟫⬜⬜⬜ | 70% |
| DSA and Problem Solving | 🟥🟥🟥🟥🟥🟥🟥🟥🟥⬜ | 90% |
| Certification | Provider | Credential |
|---|---|---|
| NLP with Classification and Vector Spaces | DeepLearning.AI | View Credential |
| Develop Generative AI Applications: Get Started | IBM | View Credential |
| Computer Vision | Kaggle Learn | View Credential |
| 160 Days of DSA | GeeksforGeeks | View Credential |
| ✅ Solved 500+ DSA problems across LeetCode and GeeksforGeeks |
| ✅ Ranked under 150 among 7.5K+ users in IIT Delhi GFG institute ranking |
| ✅ Secured 10 CGPA in Class X and received the CBSE Merit Award |
| ✅ Completed work in Cloud Computing, AI/ML, NLP, and Software Development |