I build practical, scalable AI systems with a focus on NLP, deep learning, and end-to-end ML pipelines.
Experienced in transformer models, embeddings, sequence architectures, and deploying production-ready ML applications.
I enjoy converting ideas into working ML systems — fast, clean, and reliable.
📫 Email: arjunpramod509@gmail.com
🔗 LinkedIn: https://www.linkedin.com/in/arjun-pramod
Machine Learning & NLP:
- Transformers (BERT, RoBERTa, MiniLM, BART)
- Sequence Models (RNN, LSTM, GRU)
- Tokenization, Lemmatization, NER, Sentiment Analysis, Topic Modeling
- Word Embeddings (Word2Vec, GloVe), Semantic Similarity
Frameworks & Libraries:
- PyTorch, TensorFlow, Keras, Scikit-learn
- HuggingFace, spaCy, NumPy, Pandas, Matplotlib
Tools & Deployment:
- FastAPI, Flask, Streamlit
- Docker, Git, Jupyter
- SQL, Data Pipelines
Transformers | Streamlit | BART | MiniLM | RoBERTa
- Summarizes large PDFs and answers questions in real-time
- Achieved 75% text compression, 80% QA accuracy, <3s response time
- Built using transformer-based summarization & QA pipelines
Sentence Transformers | spaCy | Streamlit
- Computes similarity between resumes & job descriptions
- Highlights matched/missing skills and gives improvement suggestions
- Built with sentence embeddings + skill extraction
CNN | Data Pipelines | Model Optimization
- Achieved 92.63% accuracy
- Designed the entire ML lifecycle: preprocessing → training → deployment
- Improved data and training pipelines for efficiency
Neural Networks | CNN
- Improved breed classification accuracy by 15%
- Developed scalable evaluation and training workflows
B.Tech — Computer Science & Engineering (AI & ML)
SRM University AP (2021–2025) • CGPA: 7.88
- 🥇 Best Paper Award, ICAIN 2025 — Lead author of “ESADN: Enhanced Spatial Attention Network for Road Accident Detection”
- 🏅 Gold Medal (Top Performer) — ProductKraft Expo 1.0
If you're interested in ML/NLP engineering, research collaborations, or AI product development — let's connect!