What are LLMs like GPT and Gemini?
OpenAI’s ChatGPT has created huge awareness around text-based generative AI. It doesn’t have a conscious understanding of what it’s writing like we do, but produces coherent text by simply predicting what the next word in a sequence might be. To do this, ChatGPT and other text generators use a type of machine learning called a large language model (LLM).
The LLM that supports ChatGPT is called GPT (Generative Pre-trained Transformer) and OpenAI release new versions of their model over time, from GPT-2 and GPT-3.5 to GPT-4, then GPT-o1.
Other notable LLMs include Gemini from Google, LLaMA from Meta, and Claude from Anthropic, but new models and enhancements are emerging all the time.
The most advanced AI assistants already use natural language processing (NLP) to recognise what your customers are saying. Now, with the integration of LLMs, we can significantly expand those capabilities to transform your customer care.
Launch a customer service chatbot in 10 minutes using LLM
AI assistants like ours have already displaced traditional chatbots because of their speed, low cost and accuracy. Using LLMs, your AI assistant can now go live in less than 10 minutes using only your website URL.
In as little as two minutes, your AI assistant is now ready to become an instant expert on your brand, business, and customer needs.
Turn your web content into a fully-functioning AI assistant
On our unique platform, AI Studio, LLMs help us pull in information from your website, or company documents you upload using RAG retrieval, and use it to automatically generate a list of question-and-answer sequences for your AI assistant to work from ― we call these flows.
Auto-generation using your own materials means the content your AI assistant uses to respond to customers is already approved for talking about your products, services, processes and offers, on your website or in marketing, sales, or customer service documents. You won’t need to do anything but check you’re happy with the wording of each flow created.
The best bit is, the responses you approve will stay the same unless you change them, so every customer asking about the same thing sees the exact same response. That way, you can keep your messaging consistent, and customers can then come to rely on accurate, up-to-date information delivered by AI.
Training suggestions to improve AI performance
Once you go live, training for your AI assistant is automated too. Every time someone asks your AI assistant a question it doesn’t already know the answer to, you’ll get a suggestion for a new flow you can add to the system. It takes an AI administrator (this can be someone on your team, or ours) less than 30 minutes a week to review automated suggestions for training to keep your AI assistant up to date and improving all the time. All the administrator needs to do is accept or reject the training suggestions and nothing goes into a flow without your approval.
Save hours of time
To train an AI assistant, you have to give it a few different examples of how a question might be phrased by different people. “I want to buy a t-shirt,” “what t-shirts are available,” and “have you got any T-shirts” are all effectively the same question. Give an AI assistant a few different examples of phrasing to learn from and it soon picks up the link. On AI Studio, LLMs can be used to generate alternate phrasing ideas instantly. An LLM like GPT-o1 has absorbed so much information from so many different places, it’ll never be short of suggestions to keep you moving through the process. You save hours of time overall not having to try and come up with countless phrasing ideas where an LLM can do it in seconds.
What can you do with LLM capability?
Businesses of all shapes and sizes are experimenting with LLM capability and using it to solve old problems in new, improved ways. We’re helping them uncover processes where use of LLMs makes business for everyone easier, faster, or more efficient. Since every company needs to pay or be paid for services, let’s take invoices as an example. There have been lots of ways over the years to submit an invoice, from sending a paper invoice in the post, to emailing a digital copy for review. Now, AI can handle it in seconds.
Chatting to an AI assistant on any channel, a supplier can now submit an invoice for payment by uploading it to the chat. An LLM can then use an established technology like OCR (optical character recognition) to extract text from the invoice and ‘read’ it to check it meets the right criteria for payment:
- The address is correct
- It has a valid date
- The VAT number is included
- The invoice number is included
- The AI assistant can even check the maths and that the amounts add up correctly
Within the chat, the AI assistant can then let the supplier know if there are errors, so they can go away and correct them before submitting the invoice again for payment.
This means there’s no need for manual checks by human admins, so they can work on higher-value tasks to help grow your business and profits, more rewarding projects they enjoy, or spend more time resolving sensitive or complicated undertakings they’re specially trained for. No time is wasted going back and forth over slow systems like email that are also prone to delay, and the whole submission and approval process happens instantly. The supplier likely gets paid faster too.
Hit the highest success rates faster with LLM support
The average success rate for the advanced AI assistants we create on AI Studio is 96%. Your AI assistant recognising a request from your customer, understanding what they need, and giving a helpful response without any input from your customer contact team is a successful chatbot interaction. Competitor platforms like Intercom have used GPT-4 to achieve a 50% success rate, but we know from experience a 99.88% success rate is achievable. Bearing in mind around 1% of questions will always be irrelevant (like someone asking a travel company what the football score is), so there’s no 100%.
Now we’re using LLMs to create AI assistants in minutes and to automate their training, you can hit a high success rate in record time, but you need an advanced platform with NLP (natural language programming), RAG, API integration, secure LLM capability, and a human in the loop all equally included to do it. Plus all the other tools and features you’ll need to grow your AI assistant over time.
On AI Studio, you don’t need to know any of these acronyms or details about AI, LLMs, GPT-4, or LLaMA to be able to use it and instantly better your customer experience.
Take advantage of all the latest LLM releases from the safety of AI Studio
Watching people experiment with ChatGPT in particular, you might be nervous about accuracy. We are too. That’s why trusted and experienced AI providers won’t be suggesting LLM content feeds directly into your live customer conversations without your approval first.
We’ve put 10 years’ worth of AI knowledge and 23 years of data engineering experience into our AI Studio platform, so you can easily create and manage your own AI assistant while taking full advantage of advanced AI. Thanks to our human in the loop system and strict guardrails, your AI assistant will operate at a safe level of accuracy from the minute it starts talking to your customers with every security and compliance responsibility already taken care of.
We’ll continue to implement new LLM releases so you can safely benefit from them without the risk that comes with experimentation, and focus all your attention on giving your customers the most sophisticated level of support.
