Deep Lessons on Deep Learning

Ashish Pujari is a principal artificial intelligence (AI)/machine learning (ML) strategist at Amazon Web Services (AWS), where he works closely with customers to solve their unique business problems.

His responsibilities include developing ML/AI roadmaps, providing technical guidance, leading discovery workshops, and collaborating with various stakeholders to confirm successful project implementation.

"I truly enjoy the hands-on nature of my work at AWS, where I get to apply my expertise in ML and AI to real-world business problems," Pujari said. "I've found that having a solid understanding of both the theoretical foundations and practical applications of ML/AI is crucial in this role."

He emphasizes that lesson to students in Northwestern Engineering's Master of Science in Machine Learning and Data Science (MLDS) program (formerly the MSiA program). Pujari is an adjunct lecturer for the program and teaches Deep Learning and Cloud Engineering for Data Science.

His approach to both courses is focusing on that combination of theory and application.

“Each course is structured to provide a solid foundation in the relevant concepts while offering practical hands-on experience through projects and labs,” Pujari said. “My emphasis lies in highlighting the significance of data preprocessing, model selection, and result interpretation. These fundamental concepts are crucial for students to grasp because they form the foundation of proficient application of machine learning techniques.” 

Those techniques are becoming more important than ever. In 2022, Fortune Business Insight valued the ML industry at $19.20 billion. It is expected to grow to $225.91 billion by 2030.

Despite an increase in conversation about AI, Pujari said there is still a misconception about it and how it relates to ML.

“They are not exactly the same thing,” he said. “AI refers to a broader concept of machines being able to perform tasks that normally require human intelligence, whereas ML is a subset of AI that focuses on enabling systems to learn from data and improve their performance without explicit programming.”

Pujari's Deep Learning course teaches about building and training deep neural networks that allow machines to learn. These networks are patterned after the structure of the human brain.

In the Cloud Engineering class, Pujari helps students explore cloud services used in data, analytics, and machine learning. Students learn how to design cloud-based solutions for industry partners while taking into consideration tradeoffs such as performance, cost, and efficiency.

In both courses, Pujari offers lessons from his 20-plus year career. Prior to AWS, Pujari held executive-level positions with CNA Insurance, consulting firm Credera, and financial/information services company Gerson Lehrman Group.

“My professional experience comes into play when sharing practical applications of concepts learned in class, as well as providing real-world examples to help illustrate complex topics,” he said. “In-class exercises are sourced from diverse sectors, enriching the learning experience with industry-specific scenarios and challenges.”

Pujari believes this approach fosters a deeper understanding of the theoretical principles he is teaching by showcasing how they are relevant and applicable as students progress professionally.

Now in his third year of teaching MLDS students, Pujari has a deep appreciation for who they are and what they can do.

“MLDS students are really driven and enthusiastic about data science,” he said. “They enjoy pushing their limits and expanding the scope of their projects. With their diverse backgrounds, they bring unique perspectives to class discussions, which makes learning more interesting and livelier.”