AI / ML Engineer • Software Engineer • Building Intelligent Systems
I am a final-year B.Tech student in Electronics & Computer Science Engineering at KIIT University, Bhubaneswar.
My work focuses on building AI systems, real-time machine learning pipelines, generative models, and scalable backend infrastructure.
I enjoy designing end-to-end intelligent systems — from model research and training to production deployment.
💡 Current Interests
- Artificial Intelligence
- Machine Learning Systems
- Generative AI
- Natural Language Processing
- Real-time AI pipelines
- Backend engineering for ML systems
- 📧 Email: deephabiswashi@gmail.com
- 💼 LinkedIn: https://www.linkedin.com/in/deep-habiswashi-54492b295/
- 🌐 GitHub: https://github.com/deephabiswashi
- 🐦 Twitter: https://twitter.com/Habiswashi_Deep
- 📸 Instagram: https://instagram.com/night_bred
- 📝 Blog: https://scorchedwithwords.blogspot.com/
- Deep Learning
- Neural Networks
- Computer Vision
- Transfer Learning
- Generative Adversarial Networks (DCGAN, CGAN, BigGAN, StarGAN)
- LangChain
- Ollama
- OpenAI Whisper
- spaCy
- Named Entity Recognition (NER)
- Retrieval Augmented Generation (RAG)
- Language Identification
- AI4Bharat IndicConformer
- NumPy
- Pandas
- Feature Engineering
- Statistical Analysis
- Model Evaluation
- Flask
- REST APIs
- AJAX
- Backend Architecture
- MongoDB
- NoSQL
- Database Design
flowchart LR
A[Live Audio Stream]
A --> B(Language Identification)
B --> C{Model Switch}
C -->|English| D[OpenAI Whisper]
C -->|Indic Languages| E[IndicConformer]
D --> F[Transcription]
E --> F
F --> G[spaCy Named Entity Recognition]
G --> H[Translation Pipeline]
H --> I[Live Captions / Output]
Tech: Python • Whisper • spaCy • LangChain • IndicConformer
- Built a real-time AI translation pipeline for live streaming platforms.
- Integrated Language Identification, ASR, and Neural Translation.
- Implemented dynamic switching between Whisper and IndicConformer models.
- Built NER-preserving translation to retain proper nouns across languages.
🔗 https://github.com/deephabiswashi/Vasha-AI
Tech: Flask • MongoDB • AJAX • Docker
- Developed full-stack MongoDB management platform with CRUD operations.
- Implemented Excel/CSV import-export for large datasets.
- Built REST API backend with Flask.
- Containerized system using Docker.
🔗 https://github.com/deephabiswashi/mongodb-manager-web-app
Tech: TensorFlow • OpenCV • GANs
- Implemented DCGAN, CGAN, BigGAN, StarGAN architectures.
- Built unified training pipelines for facial emotion synthesis.
- Trained on FER-2013 dataset (35k+ images).
Tech: LangChain • Ollama • Flask
- Built chatbot capable of answering questions from PDF, Word, and Excel files.
- Implemented vector embeddings + RAG pipeline.
Tech: TensorFlow • CNN • VGG16
- Developed EEG preprocessing pipeline using wavelet transforms.
- Built 1D CNN classifier with transfer learning.
- Generative AI
- Multilingual AI systems
- Speech Recognition
- Machine Learning Infrastructure
- Real-time AI pipelines
- LLM applications
- Google — Foundations: Data, Data, Everywhere
- IBM — Generative AI: Introduction & Applications
- IBM — What is Data Science?
- Microsoft — Python & AI Bootcamp
- Best Project Recognition — Kreativity 2.0 Project Expo (KIIT DU)
- HP Power Lab & Tata Imagination Challenge 2024
I’m always interested in collaborating on:
- AI / ML research projects
- Open-source ML systems
- Generative AI applications
- Real-time AI infrastructure
⭐ If you like my work, consider starring my repositories!