No-Code Magic: Three Top AI Chatbot Builders for Small and Medium Enterprises
Enhancing Customer Engagement: A Closer Look at No-Code AI Chatbot Tools
Hey folks, in this week's newsletter, I’m delving into the world of AI chatbots and exploring their implementation across various business scenarios, from customer service agents to knowledge management bots.
Back in March, I introduced the concept of chatbots for knowledge management, referred to then as the “corporate brain”. The system enabled employees to seamlessly access corporate data and documents to improve productivity.
In this week’s edition, I’ll recap a short history of chatbots, discuss the advantages and some disadvantages of deploying LLM-based customer service bots, and examine three distinct no-code platforms that are perfect for Small and Medium Enterprises (SMEs) who may lack the extensive tech resources of a large corporation.
Let’s dive in.
A potted history of Chatbots
You might be surprised (as I was) to discover that chatbots have a history spanning over half a century. Chatbots, often also referred to as “virtual assistants” or “chatterbots”, are surprisingly fascinating.
Below, I've highlighted some of the key milestones in the chatbot timeline that I found interesting. However, it's important to remember that countless other notable chatbots have played significant roles in advancing this technology over the decades,
1966 - E.L.I.Z.A.: Developed by Joseph Weizenbaum at MIT, E.L.I.Z.A. was one of the first chatbots ever created. It used simple pattern matching to mimic conversation but lacked any real understanding of the conversation.
1972 - PARRY: Created by psychiatrist Kenneth Colby, PARRY simulated a person with paranoid schizophrenia. It was a bit more advanced than E.L.I.Z.A. in that it had a model of the world and could exhibit emotive responses.
1980s - Racter and Jabberwacky: These were attempts to create chatbots that could mimic human-like conversation.
1995 - A.L.I.C.E. (Artificial Linguistic Internet Computer Entity): Developed by Dr. Richard Wallace, A.L.I.C.E. used an XML schema for pattern matching in conversation and was far more interactive than its predecessors.
2000 - SmarterChild: Developed by ActiveBuddy, SmarterChild was a chatbot widely used on AOL Instant Messenger and MSN Messenger. It was able to provide users with real-time information like news, weather, movie showtimes, and more.
2005 - Kuki (formerly known as Mitsuku): Created by Steve Worswick, Kuki is considered one of the world's most advanced chatbots and has won the Loebner Prize (awarded to the most "human-like" AI) multiple times.
2006 - IBM Watson: IBM's Watson represented a significant leap forward in natural language processing and was able to beat human champions on the TV quiz show Jeopardy! Watson has since been applied to various fields, such as healthcare, customer service, and more.
2010 - Apple's Siri: Siri was the first voice-activated AI assistant to hit the market on a large scale, integrated into Apple's iPhone 4S. Siri can answer questions, set reminders, send messages, and more (although it’s looking a bit dated now!).
2013 - Google Now (Now known as Google Assistant): Google's entry into the voice-activated assistant race, Google Now, could answer questions, make recommendations, and perform actions by delegating requests to a set of web services.
2014 - Microsoft's Cortana and Amazon's Alexa: Microsoft's Cortana was integrated with Windows 10, while Amazon's Alexa, released with the Amazon Echo, was designed for home automation, setting timers, playing music, and answering trivia questions.
2016 - Facebook's Messenger Bots: Facebook opened its Messenger platform to developers to create their own chatbots. These bots can provide everything from weather updates to personalized news updates and other services.
2018 - Google Duplex: An extension of Google Assistant, Google Duplex can make phone calls and carry out tasks like making restaurant reservations or booking appointments on the user's behalf, although it was shut down in 2022.
2020s - GPT-3 and beyond: OpenAI's Generative Pretrained Transformer 3 (GPT-3) and its successors represent the latest advancements in AI chatbot technology, capable of producing human-like text based on prompts.
Corporate chatbots were not a hit prior to 2022
Many of the celebrated chatbot milestones achieved above were made by major tech corporations or university research institutions.
However, chatbots aimed at the corporate market for tasks like customer service have not been so celebrated, at least not by the users of them.
Corporate chatbots, until recently, tended to be rigid, keyword or menu-driven agents, where the developer had to pre-determine the multiple different paths the user might take in advance, as well as exactly predict the words (keywords) the user would use to ask the questions.
Not surprisingly, this led to an explosion of complexity on the design side, which in turn generally led to a very poor user experience, becoming the online equivalent of navigating an automated telephone answering system.
Introducing LLM-Enabled Chatbots
However, chatbot capabilities took a quantum leap forward with the advent of LLMs, particularly with the emergence of ChatGPT in late 2022.
Chatbots enhanced with Large Language Models (LLMs) like GPT-4 are not just adept at engaging in natural, conversational interactions, but they also possess the ability to integrate external corporate data and documents through vector-based embeddings.
These AI models have the ability to autonomously reason to a degree and determine how to engage with APIs (Application Programming Interfaces) in various systems and use tools to augment and execute tasks if required.
Such advancements place us on the cusp of a revolutionary technology era fuelled by the near-human intelligence capabilities of models like GPT-4.
In the following sections of this post, I will present a high-level overview of some of the leading no/low-code platforms for building chatbots that I have recently discovered.
These platforms provide the tools that Small/Medium-sized Enterprises (SMEs) need to rapidly design, develop, and deploy their own customer service and internal knowledge management chatbots.
Let’s find out how.
Why use a chatbot in your business?
Natural language chatbots, most notably ChatGPT, and those powered by OpenAI’s GPT-3.5 and GPT-4 models have turned out to be the AI “killer app”.
In terms of technology adoption, LLM-powered chatbots quickly “crossed the chasm” into mainstream use within weeks of ChatGPT being released to the public, providing a near-perfect product-market fit.
There are a number of use cases for chatbots in a business, typically,
Customer service bots: used by customers and usually found on corporate websites (see more below), but also increasingly on messaging platforms like Instagram, Telegram and WhatsApp.
Knowledge management bots: used internally on an organisation's intranet to surface corporate information and data and help staff become more productive (the “corporate brain” idea)
Virtual assistants: used to manage calendar appointments, make bookings on the web and source information - think Alexa and Siri (that actually work!).
In the remainder of this article, I’m going to concentrate on the first use case, the Customer Service bot or agent, but the technology platforms discussed may be used in many other use cases, including those listed.
Customer Service chatbots provide a number of benefits to both the businesses that deploy them and the customers that use them. Let’s take a quick look.
Customer Service Chatbot - Benefits to Business:
Cost Savings: Chatbots can handle many simple queries simultaneously, reducing the need for a large customer service team and cutting costs significantly.
24/7 Availability: Chatbots can provide customer support around the clock without any delays, something that's usually not feasible with human-only teams.
Scalability: Unlike human agents, chatbots can easily handle spikes in query volume. This makes them ideal for handling large volumes of inquiries during peak periods or when unexpected events drive increased traffic.
Consistent Service: Chatbots provide consistent responses to customer inquiries, ensuring that the quality of service does not fluctuate.
Data Collection: Chatbots can collect and analyse customer data to identify trends, patterns, and insights that can inform business strategies.
Freeing Up Human Agents: By handling routine inquiries, chatbots free up human agents to focus on more complex tasks and provide higher-level customer service.
Customer Service Chatbot - Benefits to End Customers:
Instant Response: Chatbots can provide immediate responses to queries, reducing wait times for customers.
24/7 Support: Customers can get their questions answered at any time, even outside of traditional business hours.
Ease of Use: Many customers find text-based interaction straightforward and convenient, especially for simple queries.
Personalised Experience: Based on past interactions and data analysis, chatbots can provide personalized recommendations and support.
Consistent Interaction: Chatbots ensure consistent interaction, providing customers with the same level of service every time they interact with your business.
Customer Service Chatbot Limitations
While chatbots have many benefits, it's also important to note their limitations,
Even the most capable bots, powered by AI models like GPT-4, may lack the context to be able to service complex queries
Some customers may prefer interacting with human agents, especially for sensitive or complex issues.
LLM-based bots may suffer from hallucinations and have bias. I have outlined the main weaknesses of LLMs in more detail here.
In summary, although LLM-based bots offer a huge improvement over the previous generation of keyword and menu-based bots, a hybrid approach, combining chatbots with a hand-off to human agents, often works best.
Keep reading with a 7-day free trial
Subscribe to BotZilla AI Newsletter to keep reading this post and get 7 days of free access to the full post archives.