Muhammad Umar Ijaz

At a MedTech AI startup,My work involved cleaning and labeling multimodal audio/video datasets from psychological sessions, extracting features from facial, vocal and behavioural signals, and preparing structured, ML‑ready datasets. I also ensured data privacy through rigorous annotation and validation, and built real‑time and offline preprocessing scripts in collaboration with ML engineers to power emotion‑recognition and mood‑estimation models.

A research-driven software consultancy delivering scalable enterprise digital solutions and AI-powered products. I built automated ETL/ELT pipelines on Azure Data Factory and Databricks, cutting manual processing time by 40 % through Delta Lake. I designed Medallion Architecture pipelines that turned raw sensor data into reliable safety analytics, eliminating 90 % of data‑quality errors and deploying validation layers that delivered 90 %+ data reliability for high‑stakes decisions. I also architected scalable ingestion frameworks using Azure ML and Databricks to handle high‑volume distributed datasets and reduce reporting latency, and established CI/CD pipelines with GitHub Actions to automate code deployments and streamline development. Automated deployment workflows using GitHub Actions, integrating CI/CD patterns to streamline data processing and prepare for orchestrated scaling.

Educator | Curriculum Designer | Lesson Planner A career built around a simple belief: great learning starts with great design. I create lessons and curricula that are clear, engaging, and rooted in how students actually learn.