https://github.com/sumith25-dev/sumith25-dev
https://github.com/sumith25-dev/sumith25-dev
Last synced: 15 days ago
JSON representation
- Host: GitHub
- URL: https://github.com/sumith25-dev/sumith25-dev
- Owner: sumith25-dev
- Created: 2026-05-29T10:38:18.000Z (29 days ago)
- Default Branch: main
- Last Pushed: 2026-05-29T11:00:56.000Z (29 days ago)
- Last Synced: 2026-05-29T12:25:59.368Z (29 days ago)
- Size: 10.7 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
---
---
# ðĻâðŧ About Me
* ð B.E. in Artificial Intelligence & Machine Learning
* ðŦ BIET Davangere
* ð CGPA: **8.1 / 10**
* ðŽ Passionate about Machine Learning, Deep Learning and MLOps
* ðą Currently Learning Kubernetes, AWS and LLM Fine-Tuning
* ð Best Project Award 2026
* ðĨ 2nd Place Mini Project Exhibition
* ðž Looking for AI Engineer / ML Engineer Opportunities
---
# ð Connect With Me
## ð ïļ Tech Stack
**Languages & Core**
**Deep Learning & ML**
**MLOps & Deployment**
**Databases & Visualization**
---
# ð§ AI / ML Skills
```text
Machine Learning
Deep Learning
Data Science
Data Analytics
Computer Vision
Natural Language Processing
Large Language Models
MLOps
Model Deployment
Prompt Engineering
```
---
# ð Featured Projects
### ðĻ MLOps Image Classification Pipeline
> **Production-grade MLOps system â Built for Adobe ML Engineer Application**
[](https://ml-ops-image-classification-production.up.railway.app/docs)
[](https://github.com/sumith25-dev/ML-ops-Image-Classification)
[](https://github.com/sumith25-dev/ML-ops-Image-Classification/actions)
| Feature | Details |
|---|---|
| ðŊ Model | EfficientNet-B0 â **86.85% accuracy** |
| ðĶ Dataset | 14,000 real images (Intel Image Classification) |
| âïļ A/B Testing | 80% stable / 20% canary routing |
| ð MLflow | Model registry, versioning, experiment tracking |
| ð Drift Detection | Evidently AI â auto retraining triggers |
| ðģ Deployment | Docker + Railway (publicly live) |
| â
CI/CD | GitHub Actions â passing |
`PyTorch` `FastAPI` `MLflow` `Docker` `EfficientNet` `Evidently AI` `GitHub Actions` `Railway`
---
### ðĶī Clinical Rheumatoid Arthritis Assessment System
> ð **Best Project Award 2026 â Nirmana Exhibition, BIET Davangere**
| Feature | Details |
|---|---|
| ð§ Blood Biomarker ANN | **91.26% accuracy** |
| ðĶī Swin Transformer (X-rays) | **85.83% accuracy / 94.95% recall** |
| ðĪ AI Assistant | Groq-powered natural language queries |
| ð Validation | 5-fold cross-validation (Âą1.78% variance) |
`PyTorch` `Swin Transformer` `Streamlit` `Groq API` `SQLite`
---
### ð Customer Churn Prediction System
> Data Science Internship â Qspiders, Bengaluru
| Metric | Value |
|---|---|
| Accuracy | **92%** |
| AUC-ROC | **0.89** |
| Recall | **88%** |
| Prediction Time | weeks â **< 2 seconds** |
`XGBoost` `SHAP` `Scikit-learn` `Streamlit` `SMOTE`
---
### ð§ EchoVerse â AI Audiobook Creation Tool
> ðĨ **2nd Place â Mini Project Exhibition, JIT College 2025**
- IBM Granite-3.1 LLM + Google TTS â MP3 audiobooks
- Flask backend integrating 2 production AI APIs seamlessly
`Flask` `IBM Granite-3.1` `Google TTS` `LLM Orchestration`
# ð GitHub Analytics
---
---
# ð Contribution Graph
[](https://github.com/sumith25-dev)
---
# ðĨ Current Focus
ðđ Building Production ML Systems
ðđ Learning Kubernetes & AWS
ðđ Exploring LLM Fine-Tuning
ðđ Developing MLOps Pipelines
ðđ Open to AI/ML Engineer Opportunities
---
# ðĄ Quote
### ð Building intelligent systems that solve real-world problems.
---
