Data Engineer with 3+ years across financial services and education technology in India and Ireland. Saved analysts ~8–10 hours/week through pipeline automation. Contributed to an estimated €20,000+ in annual operational savings for an SME banking client through data lifecycle and storage migration work.
Built a deepfake detection model using EfficientNet-B0 for spatial feature extraction and 3-layer Bi-LSTM for sequential pattern recognition across compressed video frames. Preprocessed data by extracting frames, detecting/cropping faces, and analyzing noise and color distribution. Evaluated using Precision, Recall, F1-score, Confusion Matrix and ROC-AUC — achieving 90%+ accuracy on FaceForensics++ across raw and AV1-compressed formats. Applied SPSS statistical modelling to enhance detection accuracy.
Intelligent Q&A system combining LLMs with RAG over 20K+ MedQuAD medical records. Applied BERT embeddings with Pinecone vector DB for semantic search, and BM25 ranking to improve document scoring — increasing chatbot precision by 20%. Integrated LangChain and OpenAI GPT-4 for dynamic, context-aware patient responses. Deployed via Azure ML pipelines for experimentation and performance monitoring.
Scalable ETL pipeline using Azure Data Factory and Synapse Analytics to ingest, cleanse, and transform raw sales, customer, and inventory data into a structured warehouse. Designed interactive Power BI dashboards to visualize KPIs — revenue trends, regional sales, product profitability. Implemented data governance with Azure RBAC and Key Vault.
Mathematical models and optimization algorithms (linear programming, genetic algorithms) to simulate and optimize logistics and supply chain systems. Built Python-based simulations with NumPy, SciPy, and PuLP to validate model accuracy and predict system behavior under dynamic conditions — reducing manual intervention by 40%.
Actively seeking Data, AI and Analytics roles across Ireland. If you have a position, a collaboration, or just want to connect — I'd love to hear from you.