Machine Learning Engineer | Business Analytics | Data Science
I am a Machine Learning Engineer and analytics professional currently pursuing a dual Master's—an MBA in Business Analytics and an MS in Decision Analytics (Data Science) at Virginia Commonwealth University.
My work bridges machine learning with business decision-making. I specialize in developing intelligent systems that integrate NLP, computer vision, and deep learning to support high-impact, data-driven solutions. I focus particularly on RAG-based knowledge assistants, workflow automation, and the deployment of scalable AI applications.
- Natural Language Processing and Large Language Models
- Retrieval-Augmented Generation (RAG), LangChain, vector search
- Real-time computer vision systems using CNNs
- AI applications in business and decision science
- Open-source research and performance benchmarking
Multimodal Knowledge Assistant Developed an assistant for querying documents (PDF, DOCX, PPTX, XLSX) using LangChain and LlamaIndex.
Transformer-Based CV Screening Tool Built a pipeline using T5, BART, and Mistral for automated CV filtering with human-in-the-loop evaluation.
Infrared Intoxication Detection Created a CNN model for analyzing real-time infrared sensor streams, achieving 87% accuracy at 15 FPS.
Profanity Detection in Video Conferencing Implemented a hybrid LSTM + BERT model achieving 92% precision for real-time speech moderation.
Lip-Sync Generator with Voice Cloning Integrated speech-to-text and text-to-speech components to build a video generator with lip-synced avatars.
Python, R, SQL, JavaScript
PyTorch, TensorFlow, Hugging Face Transformers, LangChain, LlamaIndex
Pandas, NumPy, scikit-learn, XGBoost, Matplotlib
Docker, Flask, Django, MongoDB, Git, VS Code
- Certified Cyber Engineer – IIT Madras
- Python for Data Science, AI & Development – IBM
- Email: mondretiu@vcu.edu
- GitHub: github.com/HeavenEvolved
- LinkedIn: linkedin.com/in/ujjwal-akshith-m
I approach problem-solving through the lens of clarity, scalability, and real-world relevance. My philosophy is to build fast, validate early, and deploy responsibly—balancing innovation with practicality.