- I'm a contributor to TensorZero and the Eclipse Platform, working with Rust, Java, and backend inference logic.
- I specialize in machine learning pipelines, end-to-end ML projects, and deep learning.
- I actively use DAGsHub + MLflow for experiment tracking and best-model selection.
- Currently learning: Deep Learning, Transformers, Generative AI, LLMs, and Agentic AI.
- Databases I work with: PostgreSQL, MongoDB.
- Reach me at: aryan-work@outlook.com
- Python, Java, Rust, C++, SQL,
- NumPy, Pandas, Scikit-Learn, PyTorch, langchain,
- FastAPI
- MLflow, DAGsHub
- Git, GitHub Actions
- End-to-end pipelines (config/artifact-based pipelines)
- NLP (TF-IDF, Bag of Words, vectorizers, Word2Vec)
- Neural networks (ReLU/sigmoid/softmax, backprop from scratch)
- Model selection + reproducibility using MLOps
- Artificial Neural Nets (Feed-forward networks)
- Convolutional Neural Networks (CNN for image processing)
- RNNs (Recurrent Neural Networks)
- Variants of RNN
- LSTM (using "cell" for long-term memory)
- GRU RNNs
- Encoder-Decoder theory
- Transformer Architecture
- SQL
- PostgreSQL
- MongoDB
- Computer Vision (CNNs, ViTs, segmentation)
- Generative AI
- Langchain (in progress)
- diffusion models,
- VAEs
- LLMs
- transformers ✅
- attention ✅
- inference paths
- Agentic AI (tool-using models, planning, multi-agent systems)