Retail

Modern e-commerce offers today’s customers increased convenience. Retailers can stay relevant in a saturated market by building more personalized customer experiences and conversations with customers. As various messaging apps have grown, integrating chatbots with them is surfacing as a retail trend.  Chatbots coupled with engaging customers in a more conversational space, personalized exclusive access and previews, recommendations based on earlier purchases, best-fit suggestions for purchase are also gaining momentum to deliver an unforgettable shopping experience and service. Thus, chatbots integrated with messaging apps are becoming increasingly relevant due to their omnipresence and availability and being progressively used to provide a whole new kind of online shopping experience. 

Chatbots now supplement the role of human customer service executives, who manually carry out conversations, save received data, suggest resolutions, need individual training, technical knowledge, and proficiency, and ensure a homogenous set of responses across the entire team. An info-retrieval chatbot is more efficient than even search engines as it can guide the user to the most relevant answer instead of presenting a set of texts that might contain the correct answer. This means significant savings in costs and time and improved customer experience (CX). We have smart chatbots too, which are based on supervised and unsupervised learning, which can become intelligent and smarter by relying on AI. 

By combining the power of our flagship framework — xpresso.ai — with a decade of wholesome experience, we have ensured AI projects are delivered in a timely and robust manner, and organizations achieve a successful AI transformation journey with our systematic approach towards enterprise AI and MLOps solutions.  

Our approach to an enterprise MLOps journey begins with a robust data intake and elaborate analysis to frame the problem areas.  The second part includes applying an engineering mindset, identify enablers required for AI initiatives to be successful at enterprise scale, and preparing cognitive models. Finally, these models are deployed and managed throughout the lifecycle of the solution. Throughout the lifecycle, xpresso.ai pre-built frameworks are used to push the required transformations ahead. Coupled with a robust uptime for applications we use and near-zero chances of interruption, this means unwavering support for customers. So, what would have typically taken months to deliver can be developed and deployed in weeks, and most importantly — at a fraction of the cost.

Challenges

The utility of a chatbot can be unclear: This depends on how well the chatbot responds to the widest possible range of queries. Chatbots need to access huge volumes of data, interpret that data in real-time and produce the best-fit response to a query. The creation of this enormous dataset, refining it, creating models, and then reworking the program to constantly accept new data can be daunting.

Inability to handle queries can be a challenge: Similar queries with varying degrees of intensity or different nuances would require the chatbot to be familiar with different scenarios that sound similar and know which solution to apply to which scenario. The linear model of operation of most chatbots makes them unable to answer more than one query at a time, and worse, be unable to answer any query if faced with multiple queries at the same time.

Cultural references might be mistaken: Chatbots are often language and culture-specific. Thus, many chatbots are unable to answer standard queries if they are phrased using different languages or if they use terms specific to a different culture, even if the language in which the query is phrased is the same as the one in which the chatbot can answer. Chatbots must be regularly updated with huge amounts of data on the different ways in which a question can be asked.

Lack of a sense of conversationality with a human: Even today, many humans have an innate dislike or suspicion of anything mechanical if it can ‘talk’ or mimic other higher brain functions: one outlet for this dislike is expressing distrust of any such mechanical or digital device. Therefore, chatbots need to be designed to sound as “human” as possible, and it may be that, in the future, one of them will finally pass the Turing test.

xpresso.ai – Cognitive Solutions

• A leading Indian retail giant wanted to understand CX around its most popular line of products and services and enhance CX.
• We leveraged our client’s social media presence and developed a ‘Style Bot’ that was made available through Facebook Messenger. This bot included the following features:
  • Guided search
  • Natural language search
  • Visual search
  • Product recommendations and display (based on user input/ marketing)
  • Wish on general events
  • Coupon offering
  • Event-based promotional marketing
  • Specialist chat (chat/interaction with a human agent from the client)
• We trained the client team with the Conversational Interaction Manager and walked them through the process of how to create scripts for conversation/response messages. 
• As part of their marketing efforts, the client needed to send the links of the bot’s profile to their online users. When the bot is added as a friend or a post from the bot is liked by the online users, the bot can send out notifications to the users, thus leading them to build their personal fashion profile based on the engagements. The bot can also start wishing individuals on their friends’ list on birthdays, anniversaries and provide them instant discounts. 
• These efforts saw an increase in the Net Promoter Score to 80%.

Solution Approach

Use Case Discovery

We actively engaged with our client to capture the business requirements while observing the problem. We identified the relevant datasets and formulated a case-based approach, creating case-specific services and interfaces that would solve the business problem or improve/predict actionable insights to mitigate the problem. 

We assisted in setting up the required infrastructure – the xpressso.ai framework provides out-of-the-box development platforms. All functionality can be accessed using a Jupyter Notebook — ensuring zero-delay and plug-and-play availability of high-end hardware. Development images configured based on pre-defined templates can be installed on-premises or in a development VM within the xpresso.ai infrastructure. This enables authentication using LDAP, seamless project setup using Bitbucket, Jenkins, and Docker (ensuring build and deployment without software compatibility issues).  The project was started seamlessly with the relevant environments, which are subsequently created automatically.

The framework made available by xpresso.ai leverages the latest ML and DL tools while preparing models and includes Pachyderm-based data versioning, deployment using a Kubernetes orchestration system, Kubeflow and Spark-based ML and DL build and deployment, Istio-based service mesh enabled microservice architecture, and ELK based monitoring capability; contributing to reduction in latency time.

Data Engineering

xpresso.ai’s MLOps framework has different data adapters available through a common catalog of services that simplify interoperability and scalability concerns, enable APIs, and abstract all the technical complexities from the service consumer. This allows establishing high-end Alluxio and Presto-based rapid, inexpensive data connectivity and data collection from diverse sources (available in structured, unstructured, and streaming formats) coming in at a high velocity and in huge volumes. 

All the data sources are funneled into the data storage layer after proper validation and cleansing. The storage landscape with different storage types and extreme flexibility is built-in to manipulate, filter, select, and co-relate different data formats.

Infrastructure and MLOps Automation

The details collected, project code, data preparation workflows, and models can be easily versioned in a repository (Bitbucket, Git, etc.), and data sets can be versioned through on-premises/cloud storage. These can be added as exploratory variables by using two excellent features of xpresso.ai: 

1. Data Connectivity Marketplace libraries
2. Data Versioning

The attributes obtained are used for categorization (employing Pachyderm-based data versioning) and then performing univariate, bi-variate, and Bag of Words analysis — for both structured and unstructured datasets through xpresso Exploratory Data Analysis (Data and Statistical Analysis).  Different datasets and their different versions can be easily controlled and stored into xpresso Data Model (XDM)-enabled data store that enabled easy retrieval and storage of datasets/ files into internal XDM. 

xpresso.ai MLOps automation allows creating pipelines, train with as much data and as accurately as possible, fastest time to inference, with the ability to rapidly retrain. The xpresso Data Pipeline Management (Rapid Model Training and Experimentation) uses Kubeflow-enabled pipelines. Thus, multiple experiments using different models and datasets can be created, tested, paused, and restarted to gain better insight. 

We assisted our client’s brand custodian team in significantly improving their operational efficiency and working towards a unified CEM strategy by providing them with almost real-time actionable insights.