Amazon’s AI Product Platform Helps You Shop Smarter

Agrawal is a leading engineer at the company Amazon Web Services in Seattle. Expert in artificial intelligence and machine learningsenior member of IEEE worked on the Microsoft project Bing search engine before moving to Amazon. He also developed several features for Microsoft Teamscompany's direct messaging platform.

“I’ve been working in artificial intelligence for over 20 years,” he says. “It still amazes me how much we can do with technology.”

He shares his experience and passion for technology as an active member and volunteer IEEE Section Seattle. He organizes and conducts career development workshops that teach people how to create an artificial intelligence agent that can perform tasks autonomously with minimal human supervision.

Computer-Inspired AI Careers

Agrawal was born and raised in Chirgaon, a remote village in Uttar Pradesh. India. When he was growing up, no one in Chirgaon had a computer. His family owned a pharmacy, which Agrawal was to join after graduating from high school. Instead, his uncle and older brother advised him to go to college and find his own hobby.

He enjoyed mathematics and physics, and he decided to pursue a bachelor's degree in statistics at Allahabad University. After graduating in 1996, he received a master's degree. in Statistics, Statistical Quality Control and Operations Research V Indian Statistical Institute in Kolkata.

While at ISI, he saw a computer for the first time in a laboratory Nikhil R. PalProfessor of Electronics and Communications. Pal has been working on identifying abnormal clumps of cells in mammogram images using fuzzy c-means modela data clustering method using a machine learning algorithm.

Agrawal received his master's degree in 1998. He said he was so inspired by Pal's work that he stayed at the university to pursue a second master's degree in computer science.

After graduating in 2001, he joined Novell as a Senior Software Engineer in the Bangalore, India office. He helped develop iFoldera storage platform that allows users on different computers to back up, access, and manage their files.

Four years later, Agrawal left Novell and joined Novell. Microsoft software engineer working for a company Hyderabad Campus in India. He was part of the team developing a system to update Microsoft software with XP To Vista.

Two years later he was transferred to the development group Binga replacement for Microsoft Live Search, launched in 2006.

Microsoft Search Improvements

Live Search traffic levels were less than 2 percent, and he struggled to keep up. Google“It's a faster and more user-friendly system,” says Agrawal. He was tasked with improving search results, but Agrawal said he and his team didn't have enough user search data to train a machine learning model.

According to him, data is especially important for location-based queries, such as nearby cafes or restaurants.

To overcome these problems, the team used a deterministic approach. algorithms to create a more structured search. Such algorithms give the same answers to any query that uses the same specific terms. This process produces results by taking keywords such as places, dates and prices and finding them on web pages. To help the search engine understand what users wanted, Agrawal developed query refiners that prompted them to refine their search. The machine learning tool then ranked the results from most to least relevant.

To test new features before they launch, Agrawal and his team created an online A/B experimentation platform. Controlled tests were conducted on different versions of the products, the platform analyzed performance and user engagement metrics, and then created a scorecard showing changes in updated features.

Launched in 2009, Bing is now the second largest search engine in the world. Black Maria.

Over 10 years of working on the system, Agrawal modernized it. He also worked with the advertising department to improve Microsoft's services at Bing. Ads that match a person's search appear among the search results.

“The work seems simple,” says Agrawal, “but behind every search engine there are hundreds of engineers who provide advertising, query formulation, ranking, relevance and location.”

Testing products before launch

Agrawal was promoted to software development manager in 2010. Five years later, he was transferred to Microsoft's Seattle office. At the time, the company was rolling out new features to existing platforms without first testing them to ensure effectiveness. Instead, Agrawal said, they measured their performance after graduation, and that wreaked havoc.

He proposed using his online platform for A/B experiments across all Microsoft products, not just Bing. His supervisor approved the idea. Over the course of six months, Agrawal and his team modified the tool for company-wide use. According to him, thanks to the platform, Microsoft was able to seamlessly provide users with the latest products.

After another two years, he was promoted to senior technical manager for Microsoft Teams, which was experiencing problems with user experiencehe says.

“Many employees were receiving 50 to 100 messages a day, and it was overwhelming for them,” says Agrawal. To reduce stress, he led the team that developed the system's first machine learning feature: Trending. It identified the five most important messages that users should focus on. Agrawal also led the launch of emoji reactions, screen sharing and video calling features for Teams.

He said he was ready for a new experience in 2020 and left Microsoft to join Amazon as a development manager.

Improved shopping experience on Amazon.

Agrawal led an Amazon team that manually collected information about the company's products. retail directory for creating a glossary. The data, which included product sizes, color and manufacturer, was used to standardize language in product descriptions to ensure more consistent listings.

This is especially important when it comes to third-party sellers, he notes. Sellers listing items entered as much information as they wanted. Agrawal created a system that automatically suggests language from a glossary as the salesperson types.

He also developed an artificial intelligence algorithm that uses glossary terminology to refine search results based on what the user types into the search bar. For example, when a shopper types “red mixer,” the algorithm lists products matching the description below the search bar. The buyer can then click on an item from the list.

In 2023, the retailer's catalog became too large for Agrawal and his team to manually collect information, so they created an artificial intelligence tool that could do it for them. It became the basis for the Amazon catalog artificial intelligence system.

By collecting product information from all over the Internet, the AI ​​catalog uses large language models update Amazon listings with missing information, fix errors and rewrite product names and specifications to make them clearer for shoppers, Agrawal says.

The company expects the artificial intelligence tool to increase sales this year by US$7.5 billion, according to Fox News report in July.

Finding Purpose at IEEE

Since Agrawal joined IEEE last December, he has been promoted to senior member and has become an active volunteer.

“IEEE membership has opened doors to collaboration, mentorship and professional growth,” he says. “IEEE has strengthened my technical knowledge and leadership skills, helping me advance my career.”

Agrawal is social media Chairman of the IEEE Seattle Section. He is also Vice Chairman IEEE Computational Intelligence Society.

He co-chaired the seminar IEEE New Era AI Global Leadership Summitwhich took place from December 5 to 7 in Seattle. The event brought together government and industry leaders, as well as researchers and innovators working on artificial intelligence, smart devices, unmanned aerial vehiclesand similar technologies. They explored how new tools could be used to cybersecuritymedical field and national disaster rescue missions.

Agrawal says he stays up to date with cutting-edge technologies by reviewing 15 IEEE journals.

“An organization has a very important role to play in ensuring that everything it does is authentic,” he says. “If a journal article has the IEEE logo, you can believe it was carefully and diligently considered“