In Many Ways ...

Intelligent Agents can significantly improve the way you serve your customers.

By providing personalised customer service, recommending the best products and services to meet their needs, or helping you understand and use unstructured data.


Personalised customer service

Virtual Agents - the famed chatbots - are more than mere question and answers devices.

They are live 24 hours a day. Sure they answer questions that your customers, your employees, or your suppliers may have, but they can also place and modify orders, follow up on unpaid invoices, advise on the best products and services, all without a human being involved.

They provide personalised customer service at all levels, and can summon a human when required - they can learn from the hard questions.

It is a question of designing the right agent for the right service.

I see what you mean

Some businesses have images that mean a lot to them: designs, images, patterns, photos.

Many customers choose their purchases based on visual attributes. What if your system could offer them similar designs?

What if you could search photos based on their content? What if you could group products based on their visual characteristics?

You can. Computer Vision powers more than Facebook tags.




You like this, will you like that?

Recommender systems are commonplace in Amazon, Apple, or any larger organisation that sells products and services online. 

It is a relatively straight-forward algorithm that can be tapped in on demand by businesses of any size.

Whether the similarity of the products is based on the product content itself, on physical attributes, or even on the sound produced by your products and services, the recommender algorithm can be embedded in the ordering process, and significantly increase your sales and improve customer satisfaction.


From Start to End

From defining the best approach for your intelligent agent, collecting and analysing data, training the starting model, all the way to implementing the intelligent service, and integrating it with your existing technology.

We follow our Beyond Cognitive™ methodology. 


A typical ai project with BI GROUP

A typical AI project with BI Group involves the steps below:

1) Define the best approach: supervised learning, unsupervised learning, reinforced learning? Shall we use neural networks, regression, or search? What type? What kind of approach would be best to produce the results you require? We user our experience and your knowledge of your organisation and data to agree on the best way forward.

2) Collect data. How will we collect data to create the initial model? Will it be a one-off, or going process? Will we crawl the internet? Shall we use a big data framework, a data warehouse, the existing database, or a mere set of files? How will the training data be collected and tagged, by whom? We work with you in defining the best method and the best tools for data collection. We work with your technology of choice, or we propose our preferred one.

3) Build and Train the Initial Model. Our Data Scientists will analyse the initial data set and build an initial model that will eventually become the intelligent agent service. We can work with your technology of choice: IBM Watson, Microsoft Cognitive Services, SAP Leonardo, or Google Cloud Platform - or we can build it with our technology of choice if you prefer.

4) Build the Intelligent Agent Service. This typically involves building a web service to act as the intelligent agent. This can be delivered in an on-premise server of your choice (such as a Linux or Windows server, or virtual server), or in a cloud service of your choice: AWS, Azure, Softlayer, or as a serverless RESTful service. If you have a partner of choice, we can work with them.  

Also, what shall be the approach to keep your model learning? We design the solution and also build the ongoing learning service in this stage.

5) Integrate the Intelligent Agent with your existing solutions. An intelligent agent is simply a set of APIs. Now it is time to make it available to your existing processes. We can work with your IT team to integrate our intelligent agents into the existing solution. If you prefer, we can build the integration ourselves. Our developers work with the technology of your choice: Enterprise Business Software, such as SAP, Oracle, WorkDay, SalesForce, Microsoft Dynamics - or we can build your custom web front-end application using the javascript framework of your choice: SAP UI5, React.JS, Meteor, Angular.JS. Or maybe you would like to integrate the service into a mobile app, or build a brand new app. We develop in iOS and Android native or with React Native, Meteor, or Xamarin.


We understand that you may have your internal IT team or outsources of choice, and you have, over the years, made the right choices on the technology front. We respect that, and we are happy to operate within your environment of choice. We partner with a number of technology providers: SAP, Oracle, Google, Microsoft, IBM, AWS, DigitalOcean, Apple.

We work with the best cognitive services and APIs in the market: Watson, Leonardo, Microsoft Cognitive, and build custom models with Google's Tensorflow and Microsoft Cognitive APIs.

You don't even need a GPU. We partner with AWS, Microsoft, and IBM and can harness the power of the best cognitive servers in the market to train your intelligent agent.


Some of our enabling technologies

Don't know where to start?

Talk to us

We love what we do! We will be delighted to discuss with you how an intelligent agent could be a true catalyser for you business to provide even better customer service, and be significantly more profitable.

We are only a phone call away!

Australia: +61 (03) 9016 0506  |  USA:  +1 (650) 319 7343  |  email: