At DATAESCAPE, I designed and automated clinical data pipelines with validation, outlier handling, and continuous unit tests. I also built a delivery workflow using DVC, MLflow, FastAPI, and Docker.
For my master’s project, I processed 25.2 million e-commerce reviews on a distributed Apache Spark cluster, bringing total cluster execution time below 30 minutes. A hybrid ALS and LSH pipeline improved RMSE from 1.094 to 0.572.
I’ve also worked on distributed resource allocation and a document RAG pipeline that achieved 93.3% Recall@3. My background includes Big Data and Cloud Computing, with experience across PySpark, SQL, and cloud and MLOps tools.
