eCommerce and Conversational AI: Choosing the Right Solution

Matylda Chmielewska

eCommerce and Conversational AI: Choosing the Right Solution

With artificial intelligence being employed in more and more sectors, it is inevitable that it will also impact and reshape the retail industry as we know it.

The use of AI-enhanced tools is obviously not new for this domain - chatbots and other customer support automation solutions have been actively applied in the eCommerce industry for a while now. What’s new is how much the technology landscape has changed in this area in the last 6 months, with AI tools successfully implemented to extend what human experts can do.

Cta image

These recent changes were also the main topic of the LinkedIn event we organized on June 13, 2023. Our guest speaker was Priya M. Nair, a serial entrepreneur and founder of zwag Ai, a conversational AI-based social application. Priya was interviewed by Grzegorz Hajdukiewicz, Chief Delivery Officer at Monterail.

If you’re interested in their discussion about the intersections between eCommerce and AI, and why now is the right time to implement AI-based solutions, you can watch the recording from the event here:


You can also check out our AI series on the blog for more information about generative AI, the features ChatGPT is missing, and how to start using it in your business.

What is Conversational AI?

Conversational AI is one type of artificial intelligence - it mimics human conversations by generating responses similar to natural language and analyzing the meaning and context in real time. This technology was made possible by the use of Natural Language Processing (NLP), an AI domain that has grown exponentially in the past few years. NLP focuses on understanding and processing how humans communicate, combining it with Machine Learning for enhanced model training.

The origins of the conversational AI technology can be traced back to the 1950's and Alan Turing’s famous assumption that computers can and will be able to communicate with humans at some point in time. The very first chatbot was created not long after that, between 1964 - 1968. It was named ELIZA and it was trained to resemble how a psychotherapist would talk to their patients. It was built using the NLP core methods by Joseph Weizenbaum who was working at the Massachusetts Institute of Technology at that time.

How are chatbots different from conversational AI?

While it’s easy to see how chatbots were predecessors of modern conversational AI, the latter is not the same as this early-days technology. So, how is conversational AI different from chatbots?

Chatbots are software tools that respond to questions based on either pre-set conversation flow that mimic real-life discussions or NLPs that are able to quickly decipher the requests and respond accordingly. Previously, you had to spend weeks building possible conversation scenarios and manually typing the responses, with certain words in support requests used as triggers. Conversational AI is simply a broader term with chatbots being just one element of this technology.

AI: Potential Uses for Retail

According to Priya M. Nair, currently ‘there’s no limit to how conversational AI can be used in eCommerce and beyond’. In 2023, it’s much easier to use AI at all stages of product development, customer support management, and sales process as you don’t need to spend weeks or months developing new features and testing them with your customers. ‘Fail faster, learn faster’ has never been as viable as it is right now.

I will discuss the potential use cases for conversational AI and AI in general below, but before you start implementing any solutions, it’s important to understand what role humans can play in the process. While this technology can’t replace the need for human interaction and human specialists, it can be the extension of their capabilities, with real-life counterparts supervising the technology-based features.

Don’t forget about the human factor, and make sure that any Generative AI tools within your product are still supervised by humans.

priya-nair Priya M. Nair Founder of zwag Ai Solutions

So, where can artificial intelligence be employed in the eCommerce sector? Here’s the list of possible uses from Priya and Grzegorz.

Market research

At the product planning and market research stages, you can use Generative AI to collect unstructured data and summarize various studies and reports for you. This will allow you to validate the data and make better business decisions. With employing AI to complete these activities, your overall costs will be lower, but you’ll increase your product’s chance of success.

Instead of organizing a few rounds of customer interviews and target group discussions, you can build up a list of potential directions for your product or service using AI tools such as ChatGPT, Bard, or Claude, and then test your assumptions with human potential customers.

Sales process automation

AI-based tools can be also used to automate and enhance your sales process. By building a custom Customer Relationship Management system that uses data from your real-life conversations with your clients, you can boost your revenue and make sure that you offer your contacts services that are relevant to their needs. You can also use generative AI to summarize your calls and email conversations and add these information to the CRM.

Furthermore, AI can also serve to analyze and optimize your pipeline, suggesting which of your leads near the closing date and discovering patterns when it comes to how your clients move through the sales funnels. 

Customer recommendations

Another feature that you can build using artificial intelligence is a dynamic product/website feed that will offer your customers personalized recommendations on what other services or items they may be interested in. These recommendations can be driven by two different methods that can be used separately or combined at any stage of your customer journey. 

  • User-to-user - where what your clients see on your website is based on what users with similar interests and/or similar order history to them clicked on and bought in the past;
  • Item-to-item - where the recommendation stream comes from the identification of what other product customers may need or what items are similar to what they purchased or viewed on your website.

Support requests prioritization

Correctly identifying which support tickets your team should handle first due to their urgency, topic, or complexity can make-or-break an eCommerce business. Your customers expect your support agents’ responses to be timely and on point - and AI can help you achieve just that. With traditional chatbots, customers were often frustrated trying to get to human specialists and these tools were not always great at generating replies that were genuinely helpful. 

Modern AI tools can still rely on human-made responses with their main focus on ticket routing and making sure that your employee won’t miss any support requests that need immediate attention. Manual classification in this area can be error-prone, so in this area, it’s actually better to rely on artificial intelligence to sort through the support tickets.

Conversational AI examples

As mentioned above, conversational AI tools should always be supervised by humans and be - first and foremost - the extensions of what they can achieve. Nowadays, there are ready-made solutions that you can use to support your team and drive sales results.

AI should enable amplification of human activities - not serve as a replacement to what humans can do.

grzegorz-hajdukiewicz Grzegorz HajdukiewiczChief Delivery Officer at Monterail

So, what conversational tools can you use in the eCommerce sector? Let’s analyze a few options, including zwag Ai.


Chatfuel is a conversational AI platform that you can use to build your own chatbots and messaging tools. It’s easy to integrate with Facebook/WhatsApp API or ChatGPT, helping your team to faster respond to customer requests by handling some of these with personalized recommendations, information about your business, and more.

zwag Ai

zwag Ai is a social app that’s entirely driven by conversational AI. In addition to a feed feature that users will be familiar with, there’s also a chatbot which you can ask about local recommendations about shopping, gaming, music, travel, etc., and it will promptly deliver these to you. zwag Ai can then serve as your local search engine/social application. The solution is based on its own Large Language Model that’s developed by Priya’s team.


Heyday is a conversational AI solution developed by the team behind Hootsuite, a social media management and social listening platform. What’s important is that it’s built specifically for eCommerce, so it includes such features as product recommendations, first line of customer support, and in-chat surveys.

Sprinkl Service

Sprinkl offers a wide range of customer support and messaging tools, but there’s one of them that uses conversational AI’s capabilities to the fullest: Sprinkl Service. It allows you to redirect your support tickets to a conversational AI tool that’s able to manage the most urgent customer questions by analyzing their sentiment, context, and intent.

How to Choose the Right Tools for Your eCommerce Business?

According to Priya, the key to deciding which tool would be the most suitable for your business is mapping your processes and checking which parts of them can you actually automate. If you have numerous support requests coming in every day, perhaps AI-based ticket routing will make the most impact? Or your team struggles with regularly updating CRM with essential customer details? Then automatically summarizing the calls and adding these summaries to your CRM may be the ideal solution. 

Another direction you can take here is building a customer solution that will be based on your data and internal knowledge base using one of the available Large Language Models (LLMs) or even developing your own. 

Matylda Chmielewska avatar
Matylda Chmielewska