Muhammed Jaish K

PQAI (Product Quality AI)

• Led end-to-end development of an AI-driven system to identify root causes of defects, generating preventative rules for production scheduling.
• Developed a decision tree-based defect diagnosis system generating multiple optimal and alternative decision rules with automated visuals for end-user interpretability.
• Designed and developed a Quality Precognition system predicting stage-wise defects using Random Forest and SHAP model interpretability to isolate top defect drivers.
• Engineered machine learning (Random Forest) and deep learning (Neural Networks with Transfer Learning) models predicting steel coil mechanical properties (Yield, Tensile, Elongation), simulating process recipes to slash quality variability.

• Developed an Outlier Detection System for high-frequency sensor data utilizing statistical and machine learning models, laying foundation for patent filing.

• Developed ETL processes and pipelines using ELK Stack to enable proactive and predictive monitoring of PACS (Picture Archiving and Communication System).
• Conducted data extraction, preprocessing, and visualization to facilitate data-driven decision-making.
• Utilized Decision Tree models to predict medical image transmission latency across PACS networks.

LANGUAGES

ML / DL FRAMEWORKS

DATA SCIENCE & EXPLAINABILITY

DATA ENGINEERING & SYSTEMS

DOMAIN EXPERTISE