Machine Learning | Data Science | AI Systems
I build machine learning systems and AI-driven applications with a focus on real-world use cases.
My work spans recommendation systems, time-series forecasting, and LLM-based architectures.
Currently exploring:
- Retrieval-Augmented Generation (RAG)
- Spatio-temporal data modeling
- Applied deep learning on real-world datasets
Edit Intent — LLM Classification System
Transformer-based intent classification with fine-tuning and structured evaluation.
Movie Recommendation System
Hybrid recommendation system using collaborative filtering and latent factor models (SVD).
Deep Paper Predictor
NLP-based research recommender using TF-IDF and similarity ranking.
Energy Forecasting
Time-series modeling using LSTM, Transformer, and TCN architectures.
Multi-PDF Chatbot
RAG-based system enabling semantic search across multiple documents.
Languages
Python, Java, C++
Machine Learning
Supervised & Unsupervised Learning, Classification, Regression, Feature Engineering, Model Evaluation, Cross-Validation
Deep Learning
TensorFlow, Keras, Neural Networks (CNN, RNN, LSTM)
Natural Language Processing
Text Preprocessing, TF-IDF, Word Embeddings, Text Classification
LLM & AI Systems
Transformers, LangChain, FAISS, Retrieval-Augmented Generation (RAG)
Data Analysis & Visualization
Pandas, NumPy, Matplotlib, Seaborn
Web & Tools
Django, React
Databases
MySQL
- Designing scalable machine learning systems
- Improving model performance and evaluation strategies
- Building reliable LLM-powered applications
LinkedIn: https://www.linkedin.com/in/devanshu1013/
GitHub: https://github.com/Devanshu1013?tab=repositories