Tag Archive for: conversational ai

Human Intelligence Enables Responsible AI

By Patrick Gallagher, Managing Director, Creative Virtual APAC

The future of conversational AI is not an AI takeover. Afterall, everything artificial is in fact human made. Intelligent businesses will leverage conversational AI that is controlled by HI (human intelligence). With humans in control, conversational AI delivers much better business outcomes. These include: more sales, rich personalised experiences and improved efficiencies. Businesses and consumers can only trust the AI and by extension a brand/company if they trust the humans providing it.

Despite the scaremongering, calls for development pauses and doomsday job annihilation pronouncements AI is not going anywhere.  If anything it will become more prolific. How businesses manage this, and take advantage of the technology, will depend on who they engage with to help them navigate their AI journey and to build and deliver their AI solutions.

Generative AI and Large Language Models (LLMs)

When it comes to generative AI and LLMs, there is a need to be aware of the well documented flaws (e.g hallucinations) associated with using LLMs. However, when they are accounted for as part of the workflow process within an entire conversational AI solution it is easy to mitigate the risks. Easy, that is, if you are engaging with conversational AI experts.

With the knowledge and acceptance that AI technology out of the box is not perfect, businesses can move forward with confidence. By taking measures such as adopting best practice and having embedded safeguards when deploying AI technology within their business operations. AI that delivers value and excludes the ability to damage must be the priority.

Trust in the AI

The predicament facing businesses today is how they can build trust in AI when there is so much mistrust. Mistrust that stems from the malicious actors operating without oversight.

First and foremost, organisations should forget about do-it-yourself conversational AI solutions if they want to leverage AI to address real business challenges and create real business value. Real value is realised when AI technology is connected and integrated into the entire conversational AI ecosystem. Only then can the right solution for a business be correctly architected.

Control

Partnering with conversational AI experts who offer control of the AI with platform, as well as flexibility that can leverage the most appropriate LLM for each use case, will serve businesses best. Control over the AI, combined with controls built into the conversational AI platform, ensures that customers and employees can be confident in the integrity, honesty and accuracy of every brand engagement. This is the foundation of trust.

Flexibility and Choice

The second critical factor that organisations should consider when implementing conversational AI solutions is flexibility and choice. The continually increasing number of LLMs becoming available is giving businesses the choice they need. Businesses shouldn’t feel overwhelmed by the choice available or let it be an obstacle to decision making.

Businesses do not need to lock themselves into a particular LLM if they partner with the right partner. That is, a partner that can offer a flexible platform that natively integrates seamlessly with all LLMs. Having the widest choice of LLMs available, gives businesses the advantage of being able to leverage specifically built LLMs for verticals and/or particular use cases. This enables richer personalisation, more sophisticated conversations and greater accuracy than has even been possible previously.

Conversational AI providers

Only by working with a provider who can offer  “click and lift” functionality, so businesses can use whichever LLM they want to, with different LLMs used for different use cases, will the real value of conversational AI be realised.

LLMs and AI prevalence

If we rewind just 12 months, AI wasn’t at the forefront of mainstream conversations. The mass attention directed towards LLMs began when the world became aware of OpenAI and GPT, notably ChatGPT at the back end of 2022.

Since then, GPT version 3.5 has evolved to GPT 4.0 and we have many other LLMs that have hit the market, such as PaLM 2, Claude, Cohere, Falcon, LLaMA, Guanaco, Vicuna, MPT-30B, and they are also hitting headlines, attracting attention and being tested by businesses. The big guns in the industry, the likes of Microsoft, Google, AWS and numerous others, are all investing heavily in LLM development and in AI. The AI race is on.

Looking at the market predictions put forward by the analyst community, it is evident that AI is big business:

  • Valuates Reports says that: “The LLM market was valued at 10.5 Billion USD in 2022 and is anticipated to reach 40.8 Billion USD by 2029, witnessing a CAGR of 21.4% during the forecast period 2023-2029”
  • Markets and Research reports that “The generative AI market is projected to grow from USD 11.3 billion in 2023 to USD 51.8 billion by 2028, at a compound annual growth rate (CAGR) of 35.6% during the forecast period.”
  • QYResearch is projecting that “Global Large Language Model (LLM) market size in terms of revenue to reach 259,886.45 Million USD by 2029 from 1,302.93 Million USD in 2023, with a CAGR 141.72% during 2023-2029.”

The investment in AI and the expanding choice of LLMs available is definitely positive for businesses. Whilst trust, choice and flexibility are critical so too is the ability to leverage customised LLMs.

Realising the full potential of LLMs for business 

The guide-rails employed to control the LLMs need to be spelt out by the provider. These are inextricably linked to choice and flexibility, and sit at the core of performance, factual accuracy, reliability and also ensuring real value is realised from conversational AI implementations. The use of application specific datasets for specialised uses ensures the accuracy of results that general-purpose LLMs without the guide-rails are unable to deliver.

Looking at LLMs as a standalone technology will restrict their full potential. LLMs need to be seamlessly integrated and plugged into conversational AI solutions. LLMs must be customisable, manageable and controllable from a single platform that is built for the purpose and so fit for purpose.  They need to be customisable for specific applications as this will ensure they are more accessible and useful to every industry and more and more use cases.

It is easy for businesses to get carried away with the promises of AI and believe it is a simple off the shelf, out of the box technology that will work wonders. This is a false assumption. Businesses that do not engage with experts who understand not just the AI technology but importantly workflows, content and knowledge management and are able to provide solutions that consider the entire conversational AI ecosystem, will disappoint their customers, employees and shareholders.

Trusting the Artificial Intelligence

by Chris Ezekiel, CEO and Founder

It’s a year since ChatGPT exploded into the mainstream, grabbing headlines worldwide and bringing AI into mass consciousness. As a conversational AI company, we have been tracking and testing Artificial Intelligence (AI) models, of which there are many, for over three years. Nevertheless, when OpenAI showcased the capabilities of ChatGPT the almost overnight worldwide impact and excitement generated around large language models (LLMs) and AI, came as a bit of a surprise to many in the industry.

Working with some of the world’s largest brands, there are three key themes that dominate many of the conversations that we are having with them today. Whilst the business benefits of conversational AI are well known and accepted, and companies are embracing the use of AI, they are doing so in a considered and cautious way.

Rightly, companies leveraging LLMs are focused on: firstly, ensuring they continue to build trust with their customers; secondly, have control over the messages, answers and interactions they have with customers; and thirdly, work with providers who have extensive experience, given that their reputation, brand and future profitability are paramount considerations when implementing conversational AI solutions. As with any newish technologies there are many organisations that want to ride the wave but may not necessarily have the experience and understanding of the entire conversational AI ecosystem and how ChatGPT or other LLMs fit.

There have been many examples of LLMs hallucinating; given false, misleading or even unrelated answers. We have built controls into our conversational AI platform that provide organisations with the necessary safeguards that they need to protect their reputation whilst taking advantage of the benefits LLMs offer, and thus mitigating the potential risks. In addition to the trustworthiness of content, trust also comes into the equation with regards to a customer’s personal information.

Embedded data security and privacy guiderails for customer data is another important aspect of our conversational AI platform, enabling customer confidence as well as supporting full legislative, governance and risk compliance. The principles of designing the customer experience and ensuring there’s a strong focus on relevant, authentic, personalised content and effective knowledge management are attributes that have become even more important when implementing LLMs as part of a conversational AI solution that will deliver tangible business benefits.

Responsible AI teams are now becoming a core part of the organisation, working closely with the customer experience and compliance teams as well as marketing, sales, HR and customer service departments. Excitingly, we are in the midst of an AI revolution, and the winners will be those organisations that select experienced partners and technology platforms that have transparency and control built in from the ground up.

The changing face of Chatbots

By Maria Ward, Knowledgebase Engineer and Account Manager

We would all do well to heed the warning “with great power comes great responsibility,” when it comes to ChatGPT.  Today, industries of all sizes are making use of AI chatbots, voicebots and virtual assistants to great effect in delivering better customer and employee experiences. Consumers the world over have gotten used to and accept interacting with these ‘human-imitation’ tools to answer their questions, help them complete tasks, and guide them to resolve issues.

The face of AI chatbots, voicebots and virtual assistants is continually changing as technology advances, customer and employee demands and expectations increase, and how we communicate evolves.

Right now, there’s a lot of hype about the capabilities of ChatGPT and the large language models (LLMs) such as GPT-3 and GPT-4 that feed them.  This is not surprising given the far-reaching impact the technology can and is having across industries, professions, and social development.  People from all walks of life, from writers and developers to traders and architects, are saving time and improving efficiency by using the open AI to produce content.

The generative abilities of ChatGPT are astonishing to observe.  The capability of AI  has developed at a rapid pace over recent years, and we are at an inflection point where things are only going to develop at an ever-increasing pace!  The power of the latest LLMs is undeniable.  How this power is used practically in the design and implementation for conversational AI solutions will either add or detract from the customer and employee experience.  This is where businesses need to ensure they are speaking with experts in conversational AI.

ChatGPT in business settings

You can ask ChatGPT to give answers based on company documents, websites, and other sources of information.  When a question is asked and the answer available from the provided information source, the results are pretty reliable (but not 100%) at giving the correct information.  However, it is a recognised risk that ChatGPT will “hallucinate”, and do so very convincingly, when it cannot find the information from the source material.  This could cause significant reputational and even financial damage to a business.

So where does that leave organisations that are about to embark on their conversational AI journey and those who have existing solutions?  How do they ensure that their chatbot is powered by the latest technology and accurately represents their company?

Most organisations will not want to risk wrong information being provided to customers, however small that risk is.  Indeed, often it will be essential for public facing content to be signed off by legal departments.

Knowledge Management is king

The bottom line is that content is key when it comes to providing the best conversational AI experience. Managing the information that ‘feeds’ the chatbot is paramount for delivering the best experience to customers and employees.  It is, after all, this source information that enables accuracy of conversations.

The knowledge management system of Creative Virtual’s V-PortalTM platform allows businesses to embrace the power of LLMs. And they can do so with the confidence that the ultimate responsibility for the content is in the hands of humans.  Putting humans in control ensures that no hallucinogenic (or incorrect) answers are provided.  Businesses remain in charge of what information their chatbot can and does share.

Knowledge Management System in V-Portal

V-Portal is a platform where you can manage answers for all channels, such as web, voice, social media, and agent assist, as well as enable language capabilities, and include personalised information.  The platform supports LLMs, including native support for GPT, which enables businesses to deliver greater personalisation.  For example, using ChatGPT within the V-Portal platform you could enable a single signed off answer to be modified to change length, tone of voice (etc.) to suit the needs of the customer, based on the channel they are using.

Integration with CRM systems provides knowledge about that customer to deliver an even more appropriate personalised response.  This would be a minimal risk scenario using ChatGPT, that would enable organisations to deliver a new level of personalisation.

Content creators can be assigned specific permission levels according to their needs by granting access to only the content relevant to them. Tasks can be set against all content types giving you the power to set actions and assign these to specific users, for example for content review.

Workflows and approvals are also an integral part of the platform, allowing control over what content makes the final cut. And, with a full version history you always have visibility of what was changed and when.

A fundamental part of the platform is flow design.  This allows you to build processes to hand-hold customers through the decisions they need to make to complete their journey.

Thinking about it, V-Portal serves as the engine that brings together the great power of AI and the human intervention and judgement needed to ensure that great responsibility is maintained.

There’s more to GPT than the ChatGPT headlines

By Chris Ezekiel, Founder and CEO

AI attention over the past couple of months has been quite astounding.  Especially considering that it has been part of our every-day lives for so long already.  Face-ID to open our mobile phones – AI.  Social media – AI.  Voice Assistants like Siri and Alexa – AI.  Route mapping – AI.  Its attractiveness and almost universal application is undeniable, with global spending on AI-centric systems expected to reach $154 billion this year according to IDC.

Used extensively across every industry, delivering huge benefits and rewards, AI has been transforming businesses and the workplace for decades. In healthcare it is used in radiology results analysis and robotic assisted surgery, in banking it is the core of fraud detection, and in retail and e-commerce it is the behind personalisation, integral to effective inventory management and powers customer service chatbots.

AI used in education is enabling personalised learning and in transportation it has made driverless vehicles possible.  Businesses of all sizes are using it to deliver better customer and employee experiences with more human-like chat and voice bots, to strengthen cybersecurity, and to build more effective workflows and operations.  And these are just a few of the myriad of applications.

At Creative Virtual we have been at the forefront of helping businesses build better customer and employee experiences with innovative conversational AI solutions for nearly 20 years. Since inception we have always ensured that we practice ethical technology. We see this as a fundamental responsibility that we take seriously, and it is something that our customers expect.

There has always been adoption hesitancy when it comes to new technologies.  Issues such as job erosion, privacy, surveillance, behavioural manipulation, fake news, changing labour force/job erosion and bias are just some of the topics that come up in impact analyses of advanced technologies.

Whilst practising ethical tech comes naturally to Creative Virtual, we are not blind to the fact that there are bad actors exploiting technology and many questions on the wisdom of a concentration of power among a small elite of tech giants, with no regulatory oversight.

It is critical that the business models of companies in a position of power have robust systems in place that take account of legislative and ethical responsibilities relating to the privacy, security, discrimination and misinformation and other issues of today.

Technology has always moved faster than legislation and regulation, most recently for example we have seen this with Uber.  This is also the case in relation to AI, with questions being raised on whether self-regulation is strong enough to safeguard the rights of individuals, protect, promote and support cultural diversity, halt the spread of mis/false information, and ensure adherence to data and privacy legislation.

The continual questioning and interrogation of the social and economic impacts of new technology must happen concurrently with tech advancement and progress, with one not stopping the other.

It seems, from the media coverage at least, that ChatGPT took the world by surprise when it burst onto the scene back in November 2022.  Almost overnight, the world became obsessively captivated by AI.

The focus by mainstream media on AI technology might be new, triggered by ChatGPT, but it has been around (admittedly not as powerful or impressive – but technology is always evolving and improving) and used by businesses for quite a long time.

Nevertheless, it is being seen as ‘new’ in terms of massification and consumerisation which has led to a lot of AI hype filling print, digital and broadcast media platforms.  It has been simultaneously sensationalised, in some cases demonised, and satirised.

Most recently, it has also been politicised. We have all read that countries such as Russia, China, North Korea, Cuba, Syria, Iran and Italy have banned ChatGPT.  It should be noted that Italy is a bit of a different case to the other countries. The Italian government has banned the application because they are concerned with privacy and data, stating that “the ChatGPT robot is not respecting the legislation on personal data and does not have a system to verify the age of minor users.”

Russia, China etc. have banned ChatGPT because they either have strict rules against the use of foreign websites and applications, full on restrictions in the use of even the internet, or they have strict censorship regulations.

ChatGPT is also on the ‘watch-list’ of several countries. France, Germany, Ireland and the UK have indicated that they will be monitoring the use of the application closely for “non-compliance with data privacy laws”, and they have also raised concerns about algorithmic bias and discrimination.

It is not only governments that are questioning AI, a moratorium on the development of AI has been proposed by tech celebrities Elon Musk and Steven Wozniak.  They have put their signatures to an open letter, along with prominent and respected AI researchers such as Yoshua Bengio, Stuart Russel and Gary Marcus, asking for “… all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4.  This pause should be public and verifiable and include all key actors. If such a pause cannot be enacted quickly, governments should step in and institute a moratorium.”

Of course, the concerns voiced by governments, individuals and the industry, are valid.  They are centre stage in the ongoing debate of the ethics-morality-technology-societal quadrangle. This discussion should continue. However, proposing a pause or ban on technological advancement is not a sensible or necessary response.

As previously mentioned, Creative Virtual’s approach when building and deploying conversational AI solutions is from a position of responsibility, deploying tech in an ethical and moral way. This includes understanding the intended outcomes of the conversational AI technology solutions we build and deploy, and to debate potential unintended outcomes so we can mitigate these.

For example, it is a known that GPT-3, 3.5 and even the latest model GPT-4 are not fully reliable (humans are not reliable all the time either!) and there is no guarantee of 100% accuracy.

Yes GPT-4 is more accurate than previous versions, but it does still “hallucinate”, can give inaccurate information and harmful advice. Businesses must be able to mitigate the risks this poses to avoid financial and reputational damage and be in control of the information their company is sharing.

To ensure organisations are in control, Creative Virtual supports large language models (LLMs) including the latest versions of GPT, but we remain uncompromising on not transferring total authority to machines.

Our conversational AI solutions provide a signature blend of AI and rules-based natural language processing (NLP), with the AI component compatible with workflow functionality to allow for customisable configuration options. It also means that our systems improve continuously in a reliable way that meets the needs of an organisation.

At the same time, natural language rules can still be used to enable control over responses in instances when AI answers are insufficient. Our blended approach ensures accuracy, enables the resolution of content clashes, and delivers very precise replies when needed.

This level of enterprise-grade functionality differentiates Creative Virtual’s conversational AI platform from all others on the market today. Providing this high degree of control over the AI is critical for businesses. Organisations can be confident in the accuracy of what they communicate to customers and employees because we enable human judgement to be applied to the information created by AI.

Recent commentary about ChatGPT has highlighted examples of its imperfections as well as potential immediate and longer-term social implications. At Creative Virtual we make it possible for our customers to mitigate these risks whilst still enjoying the business benefits of large language models, specifically GPT-3.5 and 4.0 today.

Using our V-Person technology, real business concerns regarding the security, data, privacy, and accuracy aspects related to information sharing are moderated, and organisations retain full control over AI output.

We are already working with customers and introducing LLMs as part of their conversational AI solution for tasks to deliver better customer, employee and contact centre agent experiences.  After identifying specific use cases we are piloting a number of GPT capabilities that are changing the playing field, including vector matching, summarisation, text generation, translation, clustering/analysis, Q&A preparation, and using generative AI.

Implementing LLMs requires experience and expertise, especially given the rate at which AI is developing.  In conversational AI, knowledge management is critical. Creative Virtual’s orchestration platform – V-Portal – supports LLMs, enabling businesses to maximise the benefits of the latest technology and can do so safely, securely and with confidence.

Our V-Portal platform combines knowledge management with workflow management and user management, supports multiple versions of answers for a single theme which gives granular control over the responses given, and allows for optimisation for individual channels.  It also has the capability to manage multi-lingual solutions within a single knowledgebase.

Organisations using our V-Portal platform have options for presenting users with a specific response based on a variety of criteria, including channel, authenticated user profile or selected language.  The platform also supports the use of rich media such as diagrams, images and videos in addition to text and hyperlinks within answers.

The flexible architecture of V-Portal enables seamless integration of our conversational AI solutions into existing processes and technology infrastructure, ensuring business continuity and as a cloud-based solution upgrading to take advantage of the newest technologies and stay ahead of the competition could not be easier.

AI will continue to capture headlines for many years to come.  The good, bad and ugly will be debated.  How society, businesses and individuals choose to use AI is a big part of the positive impacts it will have.

As a business tool, conversational AI solutions powered by the latest in AI technology advances can supercharge employee, customer and contact centre agent experiences, whilst also delivering cost and efficiency savings, and improving productivity.

It’s all about having the right conversational AI partner who understands the technology, business challenges and responsibilities, and can build and deploy solutions that meet the real needs of business.

Contact us to find out more on how LLMs can help you deliver better employee, customer and contact centre agent experiences.

ChatGPT, GPT-3, and Your Conversational AI Solution

By Chris Ezekiel, Founder & CEO

Since the official announcement in November 2022, there has been an enormous amount of buzz and excitement about OpenAI’s ChatGPT. Industry experts are publishing articles about it, social networks are filled with comments about it, and local, national, and global news organisations are reporting stories about it. From students using ChatGPT to complete assignments for class to me getting a little help from ChatGPT to write my latest ‘Virtual Viewpoint’ column, it certainly seems like everyone is testing it out.

As a specialist within the conversational AI space, Creative Virtual is excited about what ChatGPT and the technology behind it bring to our industry. We’ve been having lots of discussions with our customers and partners, as well as internally, about how this can deliver value to businesses using our V-Person™ solutions.

ChatGPT is an extremely powerful language model that is changing quickly and will continue to get more sophisticated. However, like any deep neural network, it is a black box which is hard – if not impossible – to control. Using it as a generative tool means you can’t steer in detail what it’s going to say.  You can’t deliver reliable, accurate self-service tools if you can never be certain what response might be given.

These limitations don’t mean you should write off ChatGPT or GPT-3 (and future versions) as completely ineffective in the realm of customer service and employee support. In some cases, one might be willing to accept a certain risk in exchange for very efficiently making large chunks of information available to a chatbot. Also there are ways to use the language power of GPT in a non-generative way, as we’ll explore in this post.

In any case, ChatGPT can only ever be used as just one piece of the puzzle, like content management, integration, user interface, and quality assurance. ChatGPT alone cannot replace all of that.

One of the design features of Creative Virtual’s conversational AI platform is the flexibility to integrate with other systems and technologies, including multiple AI engines such as transformer models like GPT-3. We are currently exploring the best way to interface with this model and use it to deliver value to our customers and partners.

Let’s take a closer look at ChatGPT, how it works, and the ways it can be used to deliver customer service and employee support.

 

What kind of AI is ChatGPT and how is that different from how V-Person works?

ChatGPT is a transformer model, a neural network, and is trained to predict text continuation. It uses a variation of GPT-3 which is OpenAI’s large language model (LLM) trained on a wide range of selected texts and codes. It is extremely powerful with respect to language understanding and common world knowledge. However its knowledge is not limitless and so on its own it will not have large parts of the information needed for specific chatbot use cases. Also its world knowledge is frozen at the time it was trained – currently it doesn’t know anything about events after 2021.

V-Person uses a hybrid approach to AI using machine learning, deep neural networks, and a rule-based approach to natural language processing (NLP). The machine learning component is integrated with workflow functionality within our V-Portal™ platform so enterprises can decide the best configuration for their conversational AI tool to improve in a controlled and reliable way. At the same time, natural language rules can be used as an ‘override’ to the machine learning part to ensure accuracy, resolve content clashes, and deliver very precise responses when needed.

We developed this approach to give our customers control over the AI to create accurate, reliable chatbot and virtual agent deployments. The use of natural language rules as a fallback option to fix occasional issues and finetune responses is much more efficient than trying to tweak training data.

 

Can businesses use ChatGPT to directly answer questions from customers and employees?

At the time of writing, ChatGPT is still in a research preview stage and highly unstable with no clean API available, so it’s not possible yet for businesses to use it in this way. However with its predecessor, InstructGPT, it is. It’s also worth noting that GPT-3 is high quality only in English and a few other languages which is another potential limitation for global use.

The biggest issue with using ChatGPT to directly answer questions from customers and employees is that it does not give you control over how it will respond. It could give factually incorrect answers, give answers that don’t align with your business, or respond to topics you’d prefer to avoid within your chatbot. This could easily create legal, ethical, or branding problems for your company.

 

What about simply using ChatGPT for intent matching?

There are two ways in which GPT-3 could be used for intent matching.

The first way just uses GPT-3 embeddings and trains a fairly simple neural network for the classification task on top of that. The second option also uses GPT-3 embeddings and a simple nearest neighbour search on top of that. We are currently exploring this last option and expect to get some quality gains from that approach.

 

Can I just provide a few documents and let ChatGPT answer questions by ‘looking’ at those?

Yes, this is absolutely possible. In fact, we have offered this functionality with V-Person for several years without needing GPT but none of our clients have been interested. GPT-3 improves the quality of this in most cases, but also comes with a higher risk of being very wrong. If an organisation is interested in using GPT-3 in this way, we can support it within our platform but what we currently offer already enables us to deliver document-based question answering.

It’s important to keep in mind that using ChatGPT to answer questions from documents is only addressing one aspect of the support expected from a virtual agent. For example, no transaction triggering API will ever be called by GPT looking at a document.

 

Is it possible to give GPT-3 a few chat transcripts as examples and let it work from them?

You can provide GPT-3 with sample transcripts and tell it to mimic that chat behaviour. But unless you want a chatbot with a very narrow scope, a few transcripts won’t be enough. If there are complex dialogue flows that need to be followed, you’ll need to provide at the very least one example of each possible path – most likely you’ll need more.

This raises some difficult questions. How do you maintain those if something changes? If you try to use only real agent transcripts, how do you ensure that you have complete coverage? How do you deal with personalised conversations and performing transactions that require backend integration? It may not be too difficult to train the model to say ‘I have cancelled that order for you’ at the right time, but that doesn’t mean GPT will have actually triggered the necessary action to cancel the order.

When you really examine this approach it becomes clear that this is not an efficient way to build and maintain an enterprise-level chatbot or virtual agent. It also doesn’t address the need to have integration with backend systems to perform specific tasks. Today our customers achieve the best ROI through these integrations and personalisation.

 

What other key limitations exist with using ChatGPT to deliver customer service or employee support?

Using a generative ChatGPT-only approach to your chatbot does not give you the opportunity to create a seamless, omnichannel experience. To do that, you need to be able to integrate with other systems and technologies, such as knowledge management platforms, ticketing systems, live chat solutions, contact centre platforms, voice systems, real-time information feeds, multiple intent engines, CRMS, and messaging platforms. These integrations are what enable a connected and personalised conversational AI implementation.

With ChatGPT there is no good way to create reliable and customised conversation flows. These flows are regularly used within sophisticated conversational AI tools to guide users step-by-step through very specific processes, such as setting up a bank account. This goes a step further than just creating a conversational engagement to employing slot-filling functionality, entity extraction, and secure integrations.

You also won’t have the ability to optimise the chatbot for the channels and devices on which it will be used. This includes using rich media – such as diagrams, images, videos, hyperlinks – within answers. For example, you can’t include an image carousel to display within a messenger platform. You won’t be able to show photos or drawings to help with a new product set-up. You don’t have the ability to display clickable buttons with options for the user.

 

As ChatGPT continues to change and moves out of the research preview stage, our expert team at Creative Virtual will stay on top of new developments and opportunities this technology offers. Our mission is always to innovate in a way that will help companies tackle their real challenges and deliver real business results – and our approach to this language model is no different.

If you’re interested in discussing more about how ChatGPT and V-Person might fit with your conversational AI strategy, get in touch with our expert team here.

Hot Off the Press: The Chatbot Buyer’s Guide for 2023

By Chris Ezekiel, Founder & CEO

Chatbots and conversational AI have been gaining acceptance as essential pieces of successful customer service and employee support strategies. If your organisation doesn’t have at least one of these solutions already, it’s likely you are planning to deploy one soon or are exploring the possibility of adding one to your 2023 strategy.

Unfortunately, as adoption of this technology is increasing so is the oversaturation of the market with poor performing chatbot products. Now many live chat, CRM, and contact centre vendors are attempting to jump on the conversational AI bandwagon with their own ‘add-on bots’. This is creating both confusion for buyers and a starker divide between vendors selling add-on bots and vendors that are true conversational AI specialists.

What makes a chatbot vendor a conversational AI specialist?

A conversational AI specialist is a vendor whose core product is their conversational AI platform. It’s not just something they have tacked onto another product offering to take advantage of the growing interest in chatbots. They have an established history of delivering successful virtual agent and chatbot solutions.

A conversational AI specialist has a focus on creating successful self-service experiences through tightly integrated and personalised conversational engagement. Their technology uses artificial intelligence (AI) to improve solutions in an accurate, reliable way that gives organisations full control. They are driving conversational AI technology forward with dedicated development teams and innovative deployments.

If a CRM, live chat, or contact centre vendor you are already working with suggests adding their own chatbot to your contract, it can be tempting to quickly accept as an easy way to meet growing demand for self-service. However, it is very important that you take the time to properly evaluate their solution and how it will work for your organisation.

If you are considering the purchase of an add-on bot – or any conversational AI solution, for that matter – you should first ask yourself the following questions:

  • Am I looking for a chatbot that will deliver high rates of self-service resolution and containment to reduce my support costs and alleviate pressure on live agents?
  • Am I looking to deliver a personalised self-service experience that goes beyond just a basic FAQ bot?
  • Am I looking for a platform that fits with a composable business approach and can easily integrate with my current backend systems so I can respond to market changes with agility?
  • Am I looking for a solution that offers me scalability, flexibility, and lots of customisation?
  • Am I looking to deploy a solution that will deliver long-term self-service success?

Did you answer YES to any of these questions? Then you need the brand new Chatbot Buyer’s Guide: Benefits of Collaborating with a Conversational AI Specialist.

This comprehensive 2023 guide takes you through the key differences between a solution from a conversational AI specialist and the most common add-on bots on the market today. It also includes a quick comparison chart to help with your evaluation and purchasing decision.

Unless you’re someone working in the conversational AI space every day, it can be extremely difficult to know exactly what questions to ask when comparing chatbot products. That’s why expert guidance is so important for making a smart purchasing decision. Differences in how AI is implemented, the types of self-service experiences that can be delivered, typical pricing structures, and compatibility with composable business models are just a few of the areas covered by this guide.

Download your copy of The Chatbot Buyer’s Guide for the expert insights you need to get your 2023 conversational AI journey off to a good start. Whether you start by reading the in-depth explanations of each key chatbot capability or immediately jump to the quick comparison chart at the end, this guide will help you create an informed evaluation process for a smart purchasing decision.

If you’re interested in learning more about working with a conversational AI specialist, the experienced team at Creative Virtual is always ready to arrange a live demo and consultation session.

Wishing you Joy & Good Cheer This Season & in the New Year!

The end of 2022 is nearly here and as we reflect on the past year, the Creative Virtual team want to express our thanks to all of our blog readers, customers, and partners. We appreciate you!

The last 12 months brought lots of exciting developments for the conversational AI industry and us as a company. We reflected on some of these in our 2022 in Review blog post. Since that look back was published, one of our other blog posts – Can Conversational AI Make Your CX More Human and Empathic? – won first place in the CX Technology category of the 2022 Customer Experience Update MVP Awards!

While it’s always nice to end the year celebrating a win, we’re also busy looking forward to 2023. We have some exciting things in the works, including a new chatbot buyer’s guide and the next release of our V-Person™ technology. While you wait for those, be sure to check out our other educational resources and subscribe to our Blog – if you haven’t already, of course!

On behalf of all of us around the world at Creative Virtual, we wish you joy and good cheer this season and in the new year!

A Look Back: 2022 in Review

By Mandy Reed, Global Head of Marketing

As the end of 2022 draws near, it is time once again for Creative Virtual’s annual year in review blog post. Every year we take this opportunity to reflect on the hard work of our team, our contributions to the conversational AI industry, and a few of our company’s biggest highlights from the past 12 months.

Two of the things we are proudest of at Creative Virtual are our experienced, dedicated team and the unique expertise we provide to our customers and partners. Whether it’s through our product development or our collaborations with individual clients, it’s important to us that we consistently deliver the best solutions possible. Having an analyst group recognise us for this is always an exciting bonus – and that’s what happened again this year.

AIxOutlook conducted an independent assessment of the major conversational AI vendors in the market and found Creative Virtual to be the Innovation Excellence Leader!

“Creative Virtual is the clear Innovation Excellence Leader in a crowded and competitive conversational AI industry. Businesses collaborating with them benefit from their expert consultation, resulting in customised, integrated, and personalised solutions that deliver real business value.”

This Innovation Excellence Leadership award recognises us as the foremost conversational AI innovator driving the industry forward. You can read more about AIxOutlook’s full evaluation by downloading the report for free here. This honour means even more as we prepare for the upcoming launch of the next innovative release of our V-Person™ technology.

The analysts at Celent conducted their own evaluation of intelligent virtual assistant platforms as well, focusing in on the technology within the retail banking space. Creative Virtual was one of ten vendors included, and we are proud of our ‘Luminary’ ranking in the final report. This means we excelled in all three of the dimensions evaluated: Advanced Technology, Breadth of Functionality, and Customer Base and Support. Celent clients can access the full report here.

We were also awarded ‘Best Conversational AI Solutions’ in the SME News 2022 IT Awards. These awards recognise companies driving for innovation and focusing on client-centricity while also remaining true experts in their industry. Being recognised as the best in conversational AI made our Founder & CEO, Chris Ezekiel, contemplate what makes a virtual agent or chatbot a true conversational AI solution and share his insights in a blog post.

women leaders of conversational AIIn October I was recognised for my contribution to the conversational AI industry by being included in the Women Leaders of Conversational AI, Class of 2023! I’m honoured to have been selected to be a part of this inaugural class and am looking forward to attending the ceremony at the Project Voice Women’s Summit in April. The This Week in Voice podcast host, Bradley Metrock, dedicated an episode to introducing each of the women selected – you can listen anywhere you get your podcasts or check it out on YouTube here.

Age UK, a Creative Virtual customer since 2017, collaborated with us on a new case study exploring the four main goals they are achieving with our V-Person technology: improve discoverability of a large amount of online content; give people more ways to easily interact with and find information; resolve easy-to-answer queries online to reduce Advice Line calls; and be proactive in testing new innovations to better meet the charity’s objectives. Check out the full Age UK success story here.

In May we announced a new partnership with Service Management Group (SMG) to deliver an industry-first dynamic assistance capability. Dynamic assistance integrates V-Person conversational AI with SMG’s digital experience solution to deliver real-time support to users as they encounter issues during their online purchasing journey. Learn more about the partnership and dynamic assistance here.

2022 ALGIM conferenceWe joined our partner Enghouse Interactive in Christchurch, NZ in November for the ALGIM (Association of Local Government Information Management) 2022 Conference. Creative Virtual’s Patrick Gallagher co-presented a well-attended session on creating award-winning chatbots in local government.

Also in November, Mugdha Desai, our Head of India Operations, took part in the Agile Mumbai 2022 Conference. The event theme was ‘Artificial Intelligence for Business Agility’, and Mugdha was a featured panellist for the session titled, ‘Benefits of AI for End User’.

Founder & CEO, Chris shared his conversational AI insights through a variety of articles, podcasts, and interviews this year, including:

The Creative Virtual team continued our tradition of sharing our expertise through our annual Blog Post Celebration for Customer Service Week and CX Day in October. This year’s posts covered multi-lingual digital customer service, the members on a conversational AI team, the battle between humans and technology, and setting customer service projects up for success. You can find the whole 2022 collection here.

We aim to publish interesting and educational posts on our blog throughout the year. This year I’m proud to have two of my blog posts selected as finalists in the 2022 Customer Experience Update MVP Awards: Composable CX: Becoming Agile and Flexible in the CX Strategy category and Can Conversational AI Make Your CX More Human and Empathetic? in the CX Technology category. Voting for the MVPs – most valuable posts! – has ended, and the winners will be announced later this month.

One of the industries in which the Creative Virtual team has extensive experience is the Insurance sector. We collaborated with Insurance Thought Leadership (ITL) to produce a whitepaper exploring how conversational AI is enabling insurance companies to greatly improve their customer experience while also slashing costs. You can get your own copy of ‘The Virtual Insurance Agent’ whitepaper here.

We also put together a short, animated video to explain V-Person for Insurance, our conversational AI solution designed specifically for the insurance industry:

Another area in which we have extensive experience is improving existing chatbot and virtual agent implementations.  We published an eBook – Conversational AI Issues & Solutions: Transforming Ineffective Chatbot & Virtual Agent Projects – that explores six of the most common reasons business leaders have given for being unhappy with their conversational AI projects and ways to overcome those challenges.

2022 has been a busy and productive year for us at Creative Virtual and, as the year comes to an end, we are excitedly looking forward to 2023. We hope you’ll stay connected by subscribing to our Blog on this page and signing up here for our Monthly Newsletter.

Multi-Lingual Digital Customer Service is Easier Than Ever

By Maria Ward, Account Manager & Knowledgebase Engineer

Good day – Guten tag – Buenos días – Bonne journée – Goededag – Buona giornata – There are more than 7,000 known languages spoken in the world today. So, it’s no surprise that language is a common barrier in both personal and business interactions.

Back in 2014, the International Customer Management Institute (ICMI) published a report titled The Growing Need for Multilanguage Customer Support. Their survey of customer service leaders found that 72% said support in a customer’s native language increased their satisfaction with customer support and 58% said it increased loyalty to the brand. Over half acknowledged that offering support in a customer’s native language was a competitive differentiator.

This research is old now, but the desire of customers to have native language support is still very much there. Luckily for businesses, new technologies are making it easier for them to offer multilingual customer service on digital channels than it was in 2014.

One of these technologies is machine translation which has seen huge improvement in recent years. Developments over the past two years have greatly increased the accuracy and reliability of many translation engine applications. This has opened up new possibilities for delivering multilingual customer self-service.

For example, this year I’ve been working on several conversational AI projects with businesses taking advantage of machine translation to provide customer service in multiple languages. One is with an organisation that has used V-Person technology since 2016 on their UK website. They are an international company and became interested in exploring ways they could leverage their successful English-speaking virtual agent in other countries.

Using an automatic translation engine is a great solution for them because it is cheaper, simpler, and easier than creating a whole new virtual agent in a second language. It lets them build on the years of investment they had already made in their English-speaking virtual agent. Now they are using that same knowledgebase to provide self-service on their German website by adding translated versions of their virtual agent answers and integrating with a translation engine.

Here’s how it works: The customer enters their question in German in the virtual agent. A translation engine is utilised to translate that input into English. The translated input is then matched in the knowledgebase to the correct piece of content. The virtual agent selects the German version of the response from the knowledgebase and presents that answer to the customer.

The company started the project by identifying the top FAQs for their German website. They then provided German translations for those pieces of content. The team also worked on making any modifications to the natural language processing (NLP) to accommodate for differences in how a German user might ask those questions or ‘weird’ automatic translations that may be returned by the engine. After a successful launch of the German-speaking virtual agent, work got underway to slowly expand the content.

Another project I’ve been working on recently is for a brand-new virtual agent. One of the reasons Creative Virtual was selected as their conversational AI provider is our ability to integrate with translation engines and manage multiple languages within one knowledgebase. This company is starting their project with seven languages.

The process for this multi-lingual virtual agent has been a little different than my first example because there was no existing knowledgebase at the start. My recommendation for any organisation looking to build a new virtual agent in multiple languages is to start by finalising all content in the main language first. This will save you time with the translation work because changes to an answer typically means having to make updates to that answer across all languages.

Using automatic translation to expand a virtual agent to multiple languages is cost-effective and saves time, but it’s not a perfect solution. You’re likely to encounter content clashes and inputs that aren’t matched with your existing content. This is why you need a virtual agent management platform that has the right functionality to specifically support integration with a translation engine. The projects I’ve been working on are successful because of our V-Portal™ platform.

The right conversational AI platform will support workarounds for the content clashes and customisations for your different languages. It should also use artificial intelligence and machine learning to provide relevant ‘did you mean’ suggestions to users when their input doesn’t match with a specific piece of content. You also have the ability to set the virtual agent to ‘auto-select’ answers. This means that if the NLP fails to match the input directly with the correct answer, it pushes one of the ‘did you mean’ answers automatically as long as that answer meets a specified confidence level.

Maintenance of your multi-lingual virtual agent is also easier when you have a highly functional management platform integrated with a translation engine. When you need to make updates to an answer, you can do that quickly across all languages since all answers are listed under the same intent in the knowledgebase. Also, any changes you make to the NLP in your main language benefits all languages. And as machine translation engines improve, you automatically benefit from the most recent developments without having to do any work on your virtual agent.

The quality of your customer service affects customer loyalty, repeat business, and your brand reputation. Offering native language support can really improve your support experience. Technologies like automatic machine translation are making it easier than ever to give customers multi-lingual customer service options. Contact the experts at Creative Virtual to learn more about how we’re helping companies deliver these solutions.

Are All Members of Your Conversational AI Team Equal?

By Laura Ludmany, Knowledgebase Engineer

There is a question I came across recently which made me think and raises a good discussion for Customer Service Week: Who is the most important participant in the workflow of the development and maintenance of any AI-powered customer service tool?

Let’s imagine we build a virtual assistant from scratch for a large enterprise client where the solution must be scalable, available across multiple channels, and delivering measurable results. There are many out-of-the-box, seemingly quick solutions on the market which catch attention with claims of being up and running with little time and effort. However, these deployments are not often expandable or manageable as the real-life interaction traffic increases. These chatbots often cannot mature at the same pace as the usage, leaving a bitter taste in the users’ mouths and doing more harm than good for the organisation.

To deploy a chatbot just for the sake of having a chatbot, to tick one cool gadget off the list, to appear to be keeping up with the technology trends – none of these are good goals for a conversational AI project. The goal should be a long term one: to leverage the virtual assistant to its full capabilities; to discover new integrations, features, channels and start using it in a proactive way; to listen to your customers’ needs and feedback gathered in conversations; to broadcast news and promote products, offers, and sales to users in a centralised, accessible way.

Building and managing a virtual assistant with the goals described above, requires more people than a reader from outside the industry would probably imagine:

  • We need a salesperson to introduce the technology to the client and translate their business requirements into virtual assistant project specs.
  • We need a project manager who keeps the momentum going between the client and the team, organises the resources, streamlines the workflows, oversees the processes, and really just holds everything together.
  • We need a knowledgebase/AI engineer who designs the user journeys, builds and updates the database of the chatbot, and manages the algorithm that matches the submitted questions with the intent.
  • We need ‘hard techies’, the software engineers and developers who build the user interface, work on the different integrations, design the templates, and ultimately deploy the virtual assistant.
  • We need an analyst to look after the reporting side of the tool, understand the client’s KPIs, implement those indicators to the reporting platforms, and then deliver the required insights and statistics to the desired reporting suites.

Depending on the size and nature of the project, there can be multiple people sharing the same sets of tasks and many times there can be even more experts involved in a launch of a single chatbot.

So, then the question is: Who is the most important part in this workflow? The sales lead as he ‘brings’ the business in and has to pitch the future client? The project manager who deals with both sides and oversees everything? The AI engineers who build and maintain the ‘brain’ of the virtual assistant? The software developers who bring the chatbot to reality by building the user interface? The analytics experts who provide the reports which show the performance and measurable results of the tool?

Hint: there is no right or wrong answer. Everyone has different views and valid arguments about it. We might say very diplomatically that each and every person has equal importance in the process.

I think, based on my experience, the most important participant in a chatbot project is the client. As the conversational AI vendor, we might have the latest integrations, the coolest features on the template and the best performing chatbots ever, but our client needs to be heavily involved in the continuous journey of a conversational AI tool for real success.

There is no sadder thing for us as chatbot professionals, than to build a majestic AI tool which is then no longer looked after as it is supposed to be. There will always be new user trends evolving, new unrecognised user questions to be addressed, and new technology updates becoming available.

Hence each point of contact has a crucial role to play to win the ‘heart’ of the client, to prove and promote the value of the chatbot, to raise interest, show enthusiasm and engage with the stakeholders. Everyone in the team needs to be proactive and showcase the capabilities of the virtual assistant, whether that be through post-sales add-on integrations and launches, regular touch base meetings, analysing and improving user journeys, flagging content gaps, showing the latest technology solutions, or sharing new reporting features. We have to pass on the passion we share within our team to the client who is just starting to discover the endless possibilities and advantages conversational AI has to offer.

So, from my point of view, making the client interested, invested and an advocate for their chatbot will ultimately make them the most important participant in the chatbot workflow. As we celebrate Customer Service Week, we should recognise their crucial role. At Creative Virtual, we celebrate all our clients who are so devoted to keeping their virtual assistants successful and with whom we work hand-in-hand, day-to-day with over years and even decades.