AI/ML Engineer & Data Solutions Specialist | MSc Artificial Intelligence @ Queen Mary
Passionate about building production-ready ML systems with measurable impact. Proven expertise in:
- 🚀 End-to-end ML deployment (FastAPI, Docker, MLflow)
- 🤖 LLM optimization & prompt engineering (50+ weekly structures @ Outlier)
- 📊 Real-time analytics dashboards (30% latency reduction in Air Quality Dashboard)
- 🔍 Data-driven decision making (92% accuracy demand forecasting @ Cognizant)
Currently researching: Computer Vision architectures & efficient LLM deployment strategies
Portfolio: yvandana.github.io/Portfolio/
| Project | Tech Stack | Impact |
|---|---|---|
| Real-time Air Quality Dashboard | DuckDB, Plotly, FastAPI | 30% faster response via caching |
| Amazon Review Sentiment Analysis | PyTorch, Transformers, RoBERTa | 92% accuracy, 15% F1 improvement |
| [Gala Groceries Demand Forecast] | Python, Scikit-learn | Enabled data-driven inventory decisions |
Outlier - Data Scientist (Prompt Engineering)
✅ Optimized 50+ LLM prompt structures weekly
✅ Pioneered evaluation framework boosting consistency by 30%
Cognizant - AI Engineer Intern
✅ Built demand forecasting module (92% accuracy)
✅ Automated EDA reducing analysis time by 40%
BASE Foundation - Impact Dashboard (Volunteer)
✅ Developed SQL-backed Looker Studio dashboards
✅ Refined KPIs for cold chain impact visualization
