Dhruv Garg
> AI/ML Engineer in Training
> Computer Vision + Generative AI
I build model-first prototypes in Python and PyTorch, then harden them into full-stack AI applications with React frontends, FastAPI services, and cloud deployment workflows.
PyTorch
FastAPI
React
Python
Evolution Journey
How I am evolving
A journey from foundations to full-stack AI/ML engineering.
Foundation
predictive_ml.init()
Built confidence with predictive ML — house price prediction, training loops, and disciplined evaluation.
Vision Shift
import cv2; import torch
Moved into deeper computer vision — depth estimation, architecture decisions, and representation-focused work.
Creative Modeling
generator.forward(z)
Explored GAN-based image generation — training dynamics, instability, and the balance between quality and controllability.
Product Mindset
app.deploy(model)
Built and deployed Neural Canvas — fast style transfer in PyTorch, shipped to Hugging Face as a real-time creative tool.
Systems Growth
docker build .
Now investing in full-stack development, cloud deployment, and DevOps workflows to build complete AI systems.
Foundation
predictive_ml.init()
Built confidence with predictive ML — house price prediction, training loops, and disciplined evaluation.
Vision Shift
import cv2; import torch
Moved into deeper computer vision — depth estimation, architecture decisions, and representation-focused work.
Creative Modeling
generator.forward(z)
Explored GAN-based image generation — training dynamics, instability, and controllability.
Product Mindset
app.deploy(model)
Built and deployed Neural Canvas — fast style transfer in PyTorch, shipped to Hugging Face.
Systems Growth
docker build .
Now investing in full-stack development, cloud deployment, and DevOps workflows.
Selected Works
Projects
From model-first prototypes to full-stack AI applications — here's what I've built.
Core
Capabilities
Applied strengths across model development, reproducibility, and deployment-aware engineering.
Model Building Discipline
Iterating on architectures with clear hypotheses and measurable outcomes.
PyTorch
Deep learning framework for robust model training
CNN / Transformer
Convolutional and attention-based architectures
Depth Estimation
Multi-modal modeling and representation learning
Loss Debugging
Diagnosing training dynamics and convergence
Creative + Applied AI
Projects where engineering rigor and creative output meet.
Style Transfer
Real-time neural style transfer pipelines
GAN Experimentation
Training dynamics, instability, and controllability
Perceptual Loss
Feature-level loss functions for visual quality
Demo Deployment
Interactive inference via Gradio and Hugging Face
Systems Expansion
Moving from model-first work to complete product delivery.
Modular Python
Clean package design and reusable components
API / CLI Design
RESTful endpoints and command-line interfaces
Containerization
Docker-based development and deployment
Cloud & DevOps
Learning cloud workflows and CI/CD fundamentals
Technology Stack
Tools and technologies I work with daily
AI / ML & Data Science
Python
PyTorch
TensorFlow
NumPy
Pandas
Scikit-Learn
Matplotlib
OpenCV
Jupyter
Infrastructure & DevOps
FastAPI
Docker
Git
GitHub
VS Code
Languages & Web
JavaScript
React
HTML5
CSS3
Java
C
> Collaborate
Looking for meaningful AI internships and research work.
If your team values consistency, curiosity, and practical ML execution, I would love to contribute. Let's build something intelligent.