Portfolio | LinkedIn | DetaLabs AI
I'm an AI Engineer and Lead Data Scientist with deep expertise in building end-to-end intelligent systems from research and experimentation to production-grade deployment. I specialize in Large Language Models (LLMs), NLP pipelines, ML engineering, and AI-driven product development.
Currently the Founder & AI Lead at DetaLabs AI, where I architect and ship AI-powered solutions for real-world business challenges.
AI Engineering & LLMs
- LLM application development (OpenAI, Anthropic Claude, Gemini, open-source models)
- Retrieval-Augmented Generation (RAG), prompt engineering, fine-tuning
- AI agent frameworks: LangChain, LangGraph, AutoGen
- Model serving, optimization, and production deployment
Data Science & Machine Learning
- End-to-end ML pipeline design (data ingestion, feature engineering, training, evaluation, deployment)
- Supervised & unsupervised learning, ensemble methods, neural networks
- NLP: text classification, NER, sentiment analysis, topic modeling, embeddings
- Recommendation systems, time-series forecasting, anomaly detection
MLOps & Data Engineering
- ML pipeline orchestration and automation
- Vector databases (Pinecone, Weaviate, ChromaDB)
- Model monitoring, A/B testing, experiment tracking (MLflow)
- Cloud: AWS, GCP | Containerization: Docker, CI/CD
Programming & Tools
- Python (primary), R, TypeScript/JavaScript
- Frameworks: scikit-learn, TensorFlow, PyTorch, HuggingFace, Pandas, NumPy
- Databases: MongoDB, PostgreSQL, vector stores
- Visualization: Matplotlib, Seaborn, Plotly, Power BI
| Project | Description | Tech Stack |
|---|---|---|
| ML-Projects | Applied ML: churn prediction, NLP with NLTK | Python, Jupyter, scikit-learn, NLTK |
| Data_Analysis | EDA: Monte Carlo simulation & movie recommendations | Python, Jupyter, Pandas |
| Recommendation-Systems | Collaborative & content-based recommendation engines | Python, Jupyter |
| Question-Assitance | AI-powered question answering system | Python, OpenAI, RAG |
| End to End ML Pipeline | ML pipeline automation and orchestration | Python,mlops,machine-learning, airflow, mlflow,docker, kubernetes |
| Agentic AI PoC | Graph-based forecasting and marketing mix modeling PoC built with LangGraph, FastAPI, and deterministic ML pipelines. | Python, Langgraph,fastapi,Open AI, Claude |
| User-Entity-Resolution-Deduplication | User entity resolution and deduplication pipeline | Python, Spark, Autoencoder,Splink |
- Building production LLM applications with agentic workflows using LangGraph
- Exploring multimodal AI and advanced RAG architectures
- Open to collaborating on AI/ML research and open-source data science projects
- Ask me about LLMs, NLP, ML pipelines, AI product development
- Website: nimantha.detalabsai.com
- LinkedIn: Nimantha Bandara
- Company: DetaLabs AI
- Location: Colombo, Sri Lanka
"Building intelligent systems that turn data into decisions."