For users navigating travel e-commerce websites, the process of researching products and making a purchase often results in intricate browsing patterns that span numerous sessions over an extended period of time. The resulting clickstream data chronicle these user journeys and present valuable opportunities to derive insights that can significantly enhance personalized recommendations. We introduce TRACE, a novel transformer-based approach tailored to generate rich user embeddings from live multi-session clickstreams for real-time recommendation applications. Prior works largely focus on single-session product sequences, whereas TRACE leverages site-wide page view sequences spanning multiple user sessions to model long-term engagement. Employing a multi-task learning framework, TRACE captures comprehensive user preferences and intents distilled into low-dimensional representations. We demonstrate TRACE's superior performance over vanilla transformer and LLM-style architectures through extensive experiments on a large-scale travel e-commerce dataset of real user journeys, where the challenges of long page-histories and sparse targets are particularly prevalent. Visualizations of the learned embeddings reveal meaningful clusters corresponding to latent user states and behaviors, highlighting TRACE's potential to enhance recommendation systems by capturing nuanced user interactions and preferences

We introduce WebApp1K, a practical code-generation benchmark to measure LLM ability to develop web apps. This benchmark aims to calibrate LLM output and aid the models to progressively improve code correctness and functionality. The benchmark is lightweight and easy to run. We present the initial version of WebApp1K, and share our findings of running the benchmark against the latest frontier LLMs. First, open source LLMs deliver impressive performance, closely trailing behind GPT-4o and Claude 3.5. Second, model size has strong correlation with code correctness. Third, no prompting techniques have been found to lift performance either universally to all models, or significantly to a single model.

Large language models (LLMs) have shown impressive capabilities in natural language understanding and generation. Their potential for deeper user understanding and improved personalized user experience on recommendation platforms is, however, largely untapped. This paper aims to address this gap. Recommender systems today capture users' interests through encoding their historical activities on the platforms. The generated user representations are hard to examine or interpret. On the other hand, if we were to ask people about interests they pursue in their life, they might talk about their hobbies, like I just started learning the ukulele, or their relaxation routines, e.g., I like to watch Saturday Night Live, or I want to plant a vertical garden. We argue, and demonstrate through extensive experiments, that LLMs as foundation models can reason through user activities, and describe their interests in nuanced and interesting ways, similar to how a human would. We define interest journeys as the persistent and overarching user interests, in other words, the non-transient ones. These are the interests that we believe will benefit most from the nuanced and personalized descriptions. We introduce a framework in which we first perform personalized extraction of interest journeys, and then summarize the extracted journeys via LLMs, using techniques like few-shot prompting, prompt-tuning and fine-tuning. Together, our results in prompting LLMs to name extracted user journeys in a large-scale industrial platform demonstrate great potential of these models in providing deeper, more interpretable, and controllable user understanding. We believe LLM powered user understanding can be a stepping stone to entirely new user experiences on recommendation platforms that are journey-aware, assistive, and enabling frictionless conversation down the line.

If one was to explain ecommerce personalization in a simple manner, this could be it. When you walk into a physical brick and mortar store, what really impresses you? The ways in which the store engages you and treats you like you are their most special customer, isn't it? If you get what you are looking for, easily and quickly, without searching for it, then certainly the shopping experience is a breeze. If the owner is able to understand your preferences and shows you products according to your likes and thus, helps you save on time and effort, wouldn't you like to visit the store again and again?

Sure at first it was just the repetitive jobs that got replaced by AI. Then GANs started generating everything. Who needs a designer when a computer can put out 1000 designs a second? Obviously, I wouldn't be talking about this if I thought this was a problem. Everyone knows of deep blue, the chess application that first got everyone's attention by beating the best chess player. Since then AI has beat the best Go player and can beat anyone at competitive video games.

Technology nerds and enthusiasts have always dreamed of having a conversation with an artificial intelligence or AI. The living embodiment of the perfect AI would be JARVIS from the Iron Man movies. Just your voice, to have a conversation with your virtual personal assistant to do work for you. But that is science fiction. AI is still in its infancy and it has a long way to go to reach maturity to beat the Turing test.

Anna, Senior Consultant at M-Brain Toronto, is responsible for new business development in California as well as managing and executing strategic analysis and advisory projects. Irida, Consultant at M-Brain Toronto, manages and executes strategic analysis and advisory projects. Research by M-Brain suggests that process optimization is the most popular application of machine learning among companies across verticals including financial services, high tech, retail, manufacturing, and healthcare. M-Brain identified and analyzed 60 machine learning application cases published by vendors of machine learning tools to understand the use and application of machine learning by industry. The analysis focused on five market verticals: Financial service, high tech, retail, manufacturing, and healthcare.

A humanlike chat assistant like J.A.R.V.I.S from "Iron Man" might still be some way off, but artificial intelligence is going mainstream and the digital realm is where the impact of AI will be felt first. We are already seeing digital marketing use machine learning for granular targeting. IBM's Watson is a pertinent example. Companies like ReFUEL4 are using AI for predictive analytics to help designers figure out the best design for a campaign. Instead, it is going to be a collaborative effort.

At our recent bots, ai & messaging meetup we talked through our thoughts on simplifying the bot strategy and design process -- from objectives to conversation trees and beyond. Here is a roundup of our latest thoughts on designing and building effective messaging applications, bots and more. The first step in the bot planning process is deciding what the bot will do for the bot's creator, and what it will do for the bot's users. There's no point in in making something no-one will want to use, and no point in making something that does nothing for your business. So define both sets of objectives clearly and look for how these could overlap within a messaging application.