The World’s First Ever On Premise AI.
Using Your Own LLM.
On An Appliance.
In Your Office Network.
NOT In the Cloud.
- Keep your company data secure
- Access, retrieve and augment your own documents
- Control costs
- Create automations
- Create your own Large Language Models (LLM’s)
- Vast numbers of applications to choose from
Artificial Intelligence is going to change the World. Make sure it doesn’t ruin your business. keep control of your own data and processes before your company’s IP is leaked to the cloud.
“AI will have a bigger impact on mankind than the Industrial Revolution and the Internet combined”.
Oliver Dowden CBE Sec. State UK Government ’23 to ’24
If you own a knowledge based company. If you have business secrets and confidential information that you don’t want to be fed into the cloud as “training data” you need to act now.
You know you must evolve and you must enhance the power of AI. But if you allow your staff to upload sensitive documents into cloud-based AI models, you will lose control of your own information. Act now before it is too late.
Let Us Be Your Guide
The World of AI is made complicated by others. Let us simplify it and make your journey easier.
Founded in 2020, Hexacode delivers cutting-edge AI, blockchain, and other innovative technology solutions through a dedicated team of 15 experts. We serve both local and global markets through offices in UK and Indonesia and through our partners such as Zway in London, delivering excellence in every project.
Hundreds of Flexible Use Cases – Easy to Use. No Code Required.
Hexacubes have already been in use for over a year with Big Pharma companies, Estate Agents, Hospitality, Retail. Technology, Healthcare and Manufacturing.
Ask us to learn more about the built-in applications designed to work in your industry out of the box.
Getting started is simple as 1 – 2 -3.
- You Adopt Hexacube as a server on your own network
- You import your own data and data structures
- You instruct the built-in AI Applications what to do with your information and restrict who can see and access information from inside and outside of your own network.
Don’t Pay To Access Your Own Information
With Open AI, you pay for “tokens” that pay for the services to read your information and provide generated content based on your instructions. For example: The word “Sandwich” would be broken down into two tokens “Sand” and “wich” and charged accordingly. With OpenAI on cloud models you pay for tokens to put information in and to extract information.
Costs can get out of hand very quickly.
With your own in-house AI appliance you don’t pay extra to use AI applications. You are able to control costs and control who uses your data.
There are Hundreds of Large Language Models and Natural Language Processing Models Available
- We choose the best ones for the right applications
- We train the AI model to your own data for you
- We deliver your own appliance, trained and ready to go.
Listen to this podcast to hear about how AI in a Box is news!
FAQs
What do I get from Hexacode?
Lots of free resources, assistance, ideas and help.
In depth training courses for Leadership and Technical employees to help them navigate the World of AI.
A team of dedicated Consultants and Support Assistants committed to your success.
An AI box of your own capable of storing all of your company information and making that information available through any AI application using the best LLM for the application required.
Why should I use in-house AI instead of in the cloud?
In the end, it comes down to CONTROL.
Yes, you can trust your staff not to upload sensitive information to cloud AI engines that use your information and give it to others.
Yes, you could take a “Gung Ho” approach to costs and hope that AI usage expenses do not escalate out of control.
Yes, you could just allow anyone to develop their own business processes and hope your customers like them.
But wouldn’t it be safer and easier to have this in-house where you can control it? At least until you know what you are doing!
What are the specs of the Hexacube I buy?
Each Hexacube is built to specifications suitable for the applications you want to run. Our minimum specification will offer a RTX 4080 GPU (Ideal to run an LLM). RAM and VRAM are flexible and specified to order. Our base model Hexacube supports between 10-15 users and 15-20 use cases.
An eight-billion-parameter model (typical RAG Application) For example, will offer performance of about 60 tokens per second.
For Enterprises with the need to run AI applications on their own hardware, Hexacode will deploy the whole technology stack onto any servers with the correct specification in your own data centre or even in a 3rd party hosted data centre.
What is the support and Maintenance arrangement of the Hexacube?
Zway in London offers personalised support and maintenance agreements and SLA’s for each client in UK.
Why do you make the Hexacube appliance as a cube and not like a normal server?
The Hexacube was built to look good in your office or in your home. The future of AI involves the use of “Agents”. An agent will be similar to your “Alexa” in your home. The difference being:
Your Hexacube agent will be able to act on your behalf to undertake complex projects or tasks. Your agent may need to interact with other agents to get the job done. But your Hexacube will be able to get the job done without complicated programming or confusion. Hexacube is not just another server on your network. We anticipate you may want to access it by speaking to it. Not just having it built deep into your company network and hidden away in a sever farm.
Does Hexacode provide RAG technology?
Yes. At the heart of what we do, the AI technology, known by many as RAG is what we provide.
What is RAG? Retrieval-Augmented Generation? Retrieval-Augmented Generation (RAG) is the process of optimising the output of a large language model, so it references an authoritative knowledge base (your own company information) outside of its training data sources before generating a response.
The Hexacube does this. The LLM’s we use are stuffed with your own company data. The user can gain access to that information in the form they need, in the most helpful and accurate way. RAG output can be through Chatbots or just standard text and documents.
Does the Hexacube support NLP as well as LLM?
Yes. Different tasks and applications of AI will need deployment of different solutions. Natural language processing (NLP) and large language models (LLMs) are both important tools in the field of language processing and AI, but they have different strengths and weaknesses. For example:
NLP
NLP is a type of AI that uses techniques like machine learning, rule-based systems, and deep learning to understand and generate human language. NLP is good for structured tasks like translation and information extraction, and it’s often used to improve customer experience. NLP is also a good choice to reduce the computational power required in some cases.
LLMs
LLMs are a type of machine learning model that uses deep learning and large amounts of data to generate text, answer questions, and adapt to different situations. LLMs are good for complex tasks that require a broad understanding of language, like creative writing and conversational AI. LLMs are also useful for large-scale applications, like automating customer service interactions and digital marketing. However, LLMs might not be as precise as NLP models for specific tasks.
What are the costs of the Hexacube?
The cost of an on-site AI appliance depends on the size and needs of your business:
- Small Businesses: If your company has limited applications, a small number of users, and a modest collection of documents and data, solutions designed for small businesses and home use start at just £7,000.
- Larger Companies: For companies that need their in-house AI to handle complex tasks—like generating hundreds or thousands of documents, running chatbots for enhanced service, or supporting hundreds of users—costs increase accordingly to match these advanced requirements.
Why Choose Hexacube?
When compared to leasing your own server and hosting it in Amazon or Microsoft data centres, the one-time purchase cost of a Hexacube solution often matches what you’d pay in one year for a cloud-hosted setup with similar specifications.
Plus, with Hexacode’s solution:
- There are no extra charges for AI usage or tokens.
- Long-term costs are a fraction of what you’d spend using OpenAI in a cloud-based environment.
How Zway Helps You Succeed
Our partner, Zway, offers tailored consulting, design support, and training to ensure your team makes the most of your AI solution. From initial setup to day-to-day operations, Zway provides packages that adapt to your needs, helping you integrate AI seamlessly into your business and empowering your users to thrive.
With Hexacube and Zway, you get a powerful, cost-effective AI solution that’s ready to deliver real value.
What support or consulting do you provide for roll out of solutions
Support options available are vast through our UK Partner Zway. See the link on this line to review.