Hello, I'm Maruf Ahmed

As an aspiring computer science professional, I possess a deep passion for data insights and cutting-edge web technologies. With expertise in data analytics, visualization, and machine learning, I excel in making data-driven decisions.

As a Machine Learning Engineer at Apurba Technologies Limited, I design, implement, and maintain computer vision and NLP models for applications like image and text classification, object detection, and named entity recognition. My role involves rigorous model pre-processing, training, validation, and seamless integration into production systems. I've also managed the necessary infrastructure, encompassing cloud services, computing resources, and data storage.

My experience extends to managing and maintaining the infrastructure required for ML model training and deployment, encompassing cloud services, computing resources, and data storage. This hands-on involvement with the end-to-end process of model development and deployment has equipped me with a holistic perspective on the technology stack and its implications.

Previously, as a Data Analyst, I optimized processes through Business Intelligence solutions. I contributed to a Machine Learning project for the Florida Department of Transportation, focusing on Asphalt Condition Monitoring. As an Independent Contractor, I built dynamic web pages using Vue.js, conducted A/B testing, and developed impactful data visualizations with Power BI.

Beyond the realm of software engineering, I actively pursued research opportunities during my time at Boise State University. As a Research Assistant for the Micro-Mechanics and Smart Infrastructure Research Team, I played an instrumental role in designing and implementing a prototype for data collection, meticulously cleaning and annotating the data, and training a machine-learning model using TensorFlow.

Through my endeavors, I achieved a remarkable 96% overall prediction accuracy in pavement surface crack detection, employing the state-of-the-art Single Shot Multi-Box Detector architecture. Moreover, my contribution extended to the field of material science, where I supported the Material Science for Nuclear Reactor Research Team in building a comprehensive dataset of potential molecular structures and their traits. Employing powerful tools such as Pandas, Matplotlib, and Numpy, I executed data visualization and analysis, enabling the team to gain valuable insights into materials and their behaviors.