Today, AI is everywhere: From hard-hitting headlines about the threat of AI’s unregulated spread, to artificial intelligence-generated memes. It’s impossible to ignore how quickly this tech is reshaping the world around us. And alongside growing interest in this technology, recent advancements (looking at you, ChatGPT) have seen trust in AI skyrocket.
Our in-house market research found that 92% of respondents said their trust in AI increased over the past 12 months. And 88% of business leaders reported that their customers’ attitudes toward automation improved in the same period. With 69% of customers preferring to self-serve over speaking with a company rep — more and more companies are looking to invest in customer service AI solutions.
Customers and businesses alike are getting excited about AI and its potential to make their lives easier, drive efficiency, and provide better experiences. In the support sphere, CX leaders across industries are looking to see how this tech can benefit their brand. But in a crowded market — where every AI-powered support automation provider offers the world (and more) — it can be hard to choose the right AI solution for your customer service.
To help you sort the best from the rest, we’ve put together a list of the top 10 AI solutions for customer service. But first, let’s go back to basics.
What is AI for customer service?
Until OpenAI threw their hat (read: human-like generative AI bot) into the ring, conversational AI was the industry gold standard. All the leading support automation solutions were powered by conversational AI. But since ChatGPT burst onto the scene, automated customer service software providers have been hustling to integrate the latest and greatest AI technology, generative AI, into their product offerings.
Check out our own journey to build the tech behind ChatGPT into our product.
So what exactly is conversational AI?
Essentially, conversational AI is the set of technologies that allows humans and bots to communicate with one another. If you’ve ever asked Alexa to play that song that’s been stuck in your head, or received an automated answer to an email you sent to your favorite brand about a return — these interactions are powered by conversational AI.
Conversational AI works using a combination of a few principle technologies:
- Machine learning: This is a type of technology that gathers information from each interaction to learn and improve over time
- Natural language processing (NLP): This is a form of artificial intelligence that can understand and respond to human language
With conversational AI software, a virtual agent is able to understand the meaning behind human input (via text or voice) and provide an appropriate response. But these responses have to be built out in dialogue flows by conversation designers or automation managers.
Or at least that’s how it used to work.
Enter: Generative AI
Generative AI is also a form of machine learning. But with this technology, you can generate images, music, videos, and other types of content — beyond text. Unlike traditional AI systems that recognize patterns to make predictions, generative AI (the name is a bit of a giveaway) can actually create entirely new content.
Text-based gen AI is powered by large language models (LLMs). And these models are trained on vast data sets: ChatGPT, for example, was trained on all of the written content available on the internet prior to 2022. As well as generating written output that mimics human creativity, these models can summarize text, analyze message sentiment, perform translations, and more. Pretty impressive, right?
And business leaders are taking note. The most forward-thinking brands are already looking to see how they can wield this technology to provide faster, better, more efficient customer support.
More on the power of gen AI
Business benefits an AI solution for customer service brings
Onboarding an AI solution for customer service brings a whole host of benefits — from streamlined support operations to more joyful CX. The best customer service automation software allow companies to:
- Instantly resolve customer requests, 24/7, without agent involvement
- Reduce handling times and support agents in their roles by summarizing tickets and intelligently triaging requests
- Free their support teams from repetitive, manual tasks and provide agents with opportunities for upskilling
- Simplify and speed up the bot training process by generating training data and identifying new intents to cover
- Cut costs and drive efficiency while delivering more effective support
- Build flexibility and resilience into their support departments, to handle unexpected contact spikes and scale as their business grows
And these aren’t the only benefits brands will see. Want to hear from your peers about their successes with AI automation? Check out our customer stories hub.
How to choose the right customer service AI solution for your company
When it comes to choosing the best AI automation partner for your brand, there are a couple of things it's important to keep in mind:
- Your specific business goals and the pain points you want to solve with automation: Want to deflect FAQs? Then a gen AI knowledge base (KB) bot might be your best option for getting started. Or to save on manual agent effort by routing requests to the right department? For that you’ll need a ticket automation solution.
- How high your support volumes are: For small businesses with low numbers of requests, the ROI on an advanced automation solution might not be worth it (you don’t need a bluetooth-connected power tool to knock a nail into the wall — a hammer will do).
- The sophistication of your existing support setup: Are you using CRM software and multiple back office tools? Then choosing an AI partner that can integrate into your support tech ecosystem is key.
- The channels you use to communicate with customers: If your customers contact you via social media, then you’ll want an automation solution that can cover social messaging apps. If chat is their preferred channel, you might not need a provider that can automate email tickets.
So there’s a lot to think about before even starting to shop around for providers (and if you need more support, check out this article on how to evaluate AI providers). Avoid getting caught up in the generative AI hype if it doesn’t make sense for your business right now. But if you decide that a generative AI solution for customer service is right for you — here are the top contenders.
The 10 best generative AI solutions for customer service
Finding a new technology partner can feel overwhelming. To help you choose the best customer service AI for your brand, we’ve put together a list of the top 10 providers of gen AI solutions.
Note: Product features accurate as of 5 July 2023, subject to change as per ongoing iteration.
Best for flexibility, innovation, and ongoing collaboration
UltimateGPT is the latest product iteration from our team at Ultimate. We’ve long been at the forefront of AI innovation — and our LLM-powered solution cements this position as leaders in the support automation space.
Connect your knowledge base or FAQ page to UltimateGPT and you can create a custom bot in minutes — no training, no maintenance. Instead of simply serving up links, this gen AI agent finds the correct answer, summarizes it, and instantly resolves your customers’ questions in 109 languages. This product is available standalone, or as part of a broader suite of advanced automation tools.
Our LLM offering works across chat and tickets, and you can choose from 4 bot personalities to mimic your brand identity. We’re partnered with best-in-class voice assistant provider, PolyAI, to offer truly omnichannel support coverage. As well as resolving customer requests, you’ll benefit from features that support agents in their roles (like adjusting tone of voice) and assist bot builders (such as generating training data to automate the process of bot training itself).
And the cherry on top? UltimateGPT works in tandem with our existing intent-based automation product. We call this the hybrid approach. It means you don't have to choose between the conversational experience of generative AI, and the ability to perform actions and resolve complex cases end-to-end that comes with pre-built dialogue flows. Yep, with UltimateGPT you really can have your cake and eat it too.
We’re building fast and continuously adding new features to our generative AI offering. Here’s a taste of what’s coming up next:
- Generating conversation summaries to give agents the context they need to work more effectively
- Advanced analytics that give you insight into which articles UltimateGPT use the most (and the least) to generate answers
- Identifying gaps in your knowledge base where there was no article to cover a customer's question
2. OpenAI API
Best for a bot that can discuss anything and everything
OpenAI’s API gives businesses a direct integration with the original generative AI bot: ChatGPT. You would think that to deliver the best experience for your customers, it makes sense to choose this option. But open source models are foundational technologies — which means they’re missing the application layer that provides seamless UX and includes guardrails to keep hallucinations in check.
An API integration with OpenAI is a cost-effective option — and gives your customers access to the entirety of written human knowledge. While this allows your bot to wax lyrical about any topic (from pancakes to politics) there’s a higher risk of hallucinations or incorrect answers. On top of this, latency issues can slow response times and keep your customers waiting.
3. Zendesk Advanced AI
Best for improving workflows and agent efficiency
Zendesk Advanced AI is the gen AI customer service solution from leading help desk software provider, Zendesk. Powered by OpenAI, this set of features is available as an add-on to the Zendesk suite (meaning it isn’t possible for companies to use as a stand-alone product).
Currently, most of the Advanced AI features focus on agent productivity — as opposed to customer-facing interactions. These include AI-generated customer insights, macro suggestions, intelligent triage, message sentiment analysis, and more.
4. Intercom’s Fin
Best for companies with multiple FAQ pages
Fin by Intercom is another support solution powered by OpenAI, with an application layer built on top of GPT-4. Intercom promises resolution rates of 50% — and claims to be the safest and most accurate gen AI bot on the market (but, full disclosure: Intercom doesn’t offer any data to back this up).
Marketing hyperbole aside, Intercom’s GPT product can draw on multiple data sources to generate answers and it’s customizable — so you can decide if customers will see LLM-generated answers or pre-designed dialogues. As of June 2023, Fin is only available in English, but Intercom plans to roll out their generative bot to 43 languages soon.
As well as Fin, Intercom offers an AI assist tool to help agents work more effectively. Features include conversation summarization, redrafting responses into a different tone, and knowledge base article generation.
Best for generative voice support
Looking for a generative AI support solution that covers voice as well as text? Ada could be the provider for you. Powered by OpenAI’s GPT-3 model, Ada’s platform (according to them) is the world’s first gen AI support automation solution that works for voice.
As well as offering a knowledge base bot to field FAQs, Ada lets you choose from 4 personas to keep answers in line with your brand personality — and the option to jazz up generated messages with emojis. Their solution can also generate training data to improve bot accuracy, and assist agents by automatically rephrasing responses into a more conversational tone.
Best for a simple knowledge base bot
Keeping it simple with their generative AI solution, ecommerce-specific support automation platform Zowie offers a one-minute chatbot builder (if you work in any industry beyond ecommerce, Zowie isn’t for you). All you have to do is copy and paste the URL of your FAQ page — and your gen AI chatbot is ready to go.
This is great for retail brands looking for a quick and easy solution. But to see the most value from generative AI, it’s worth going with a provider that can cover more advanced use cases (like generating training data, summarizing conversations, or integrating with action-oriented dialogue flows to automate complex queries end to end) than a KB bot.
Best for having been the first
CS automation platform Forethought has gone so far as to apply for a trademark for their gen AI product, SupportGPT. Another provider making bold claims, Forethought markets itself as the world’s first generative AI platform for customer support. But what does their LLM bot do (as of June 2023)?
On the customer-facing side, SupportGPT can answer common questions and generate suggested replies for agents — helping them respond to customers more quickly. Forethought also offers generative email automation. On the agent efficiency side of things, SupportGPT can analyze message sentiment and intent to prioritize cases. Plus, this AI solution can identify gaps in your knowledge base and generate new articles.
Best for intent-based bot training
Perhaps it’s not the first gen AI platform for voice (we’ll let them fight it out with Ada for that title) but boost.ai also supports generative voice automation, as well as chat. This AI provider has also gone for a hybrid model, combining LLMs with its existing intent-based automation solution.
While boost.ai can deploy gen AI in customer interactions, they focus on agent productivity. They highlight rewriting agent answers into a different tone and summarizing support conversations to smooth agent handover. This product also supports speedier bot training.
Best for gen AI skeptics
The generative AI support automation solution from Netomi emphasizes caution when it comes to customer-facing conversations. Their gen AI bot connects to your knowledge base, to provide natural responses to FAQs without needing to create dialogue flows. And controls exist to allow agents to review and approve these answers.
But most of Netomi’s new gen AI features focus on supporting agents in their roles. These include making bot training easier, generating sample responses that agents can edit, and analyzing conversations to identify common support topics.
Best for ecommerce brands with people-power to spare
Thankful is a generative AI support automation provider for retail and ecommerce brands. Instead of creating a new LLM-powered product, Thankful have incorporated gen AI into their existing FlowsNext experience builder.
This allows their AI support solution to connect with back office systems and perform actions. The bot can also analyze message sentiment, send conversational responses to customers, summarize notes, and more.
These upgrades to FlowsNext are great for existing Thankful customers. But for brands that are new to AI automation, this solution will take longer to set up than other gen AI product offerings — as you’ll have to manually build these flows from scratch.
Now, it’s up to you
When it comes to choosing a gen AI solution for your customer service, the most important thing to consider is whether a provider can support your specific use cases. But as you’ve seen, it can be hard to cut through hype around generative AI. So do your due diligence, book demos with different providers, and shop around before committing.