Asad Khalid

ML engineer shipping computer vision, predictive maintenance, and LLM systems to production. At Pumacy I productionized DeepLabV3+ via FastAPI for automotive VR (workflow time −80%, ~98% pixel accuracy); built a prescriptive model for metal AM (+30% print success, acted as technical liaison); deployed LSTM-based anomaly detection across turbines/batteries (unplanned downtime −25%); and led domain LLMs for nutraceuticals, architecting a RAG layer that lifted training efficiency and sped iteration.

Application of computer vision and machine learning for industrial automation. - Use of Deep learning libraries (keras, tensorflow, caffe), stereo and time-of-flight cameras, GUI designing based on Python (Tkinter). - Color and Depth based Image processing including segmentation for object detection, tracking and classification. - 3D Design using Blender

- Study and implementation of CNN, R-CNN and state of the art real-time object detectors like YOLO, SSD. - Data Acquisition and processing from time-of-flight cameras like DepthSense for object detection, tracking and classification.

- Study and operation of PLCs (Siemens) for various automation tasks, using LAD and FBD programming. - Integrating PLCs with HMI modules for human-machine interaction.

- Study and monitoring the operation of Jenbacher gas engines (type 2,3,6) using DIA.NE. Study of Cummins diesel engines and synchronous generators.

* Computer Vision * Machine Learning * Automation * Robotics

* Automation * Image processing and computer vision * Robotics