Taimoor Khan2

I’m an AI Engineer focused on applied machine learning and production-grade AI systems. At Ninth Dev Solutions, I engineered an LSTM-based predictive analytics platform over EHR data from 15 hospitals, forecasting patient health trajectories and adverse events at scale.

I build models that ship: I created a stacked ensemble (XGBoost + LSTM) for multivariate solar-energy forecasting with production deployment and automated retraining via MLflow pipelines. I also designed GRU-based sequence models over 5TB of retail transaction data to extract purchase-propensity and churn signals, and applied K-Means + PCA on hospital operations to cut operational costs by 12% and improve resource allocation by 18%.

My work also spans edge computer vision and NLP. I led end-to-end delivery of YOLOv4 edge deployment (mask detection) and BERT-based NLP tools, owning model quantization, pruning, and CI/CD integration—so performance improvements made it to production reliably.

My flagship project, CliniSynapse AI (Clinical Decision Support System), is a RAG + LLM system over 167K+ real patient case records, delivering sub-200ms end-to-end clinical query responses. I architected a multi-stage retrieval pipeline using FAISS (BioBERT embeddings), domain-aware scoring, and cross-encoder reranking, integrated live PubMed evidence retrieval with strict grounding to block hallucinations, and shipped a full-stack doctor-facing experience with caching and parallel GPU+CPU execution.