At this stage, we will discuss the needs and expectations of utilising AI for business. Concurrently, we will analyse the technical possibilities for realising the service. This will involve analysing the available data sources for training the service, as well as the internal resources that are readily available to undertake the desired project or product. We will address the following questions.

1. What are the business challenges you aim to address with AI and data?
2. What are the needs and requirements for deeming the project/data product successful?
3. How complex do the data products need to be in order to achieve your business goals?
4. Do you have available data for training purposes?
5. Do you possess the necessary tools and resources to execute the project?

Output: 

• Briefing document 
• Proposal 

During this step, we walk through the vision workshop together with your team to understand the desired functionality of the data and AI product as well as relevant use cases.  During the workshop, we will answer the following questions.

1. What are the business’ priorities in terms of solutions?
2. What are the business constraints?
3. What is your vision for the transformation of your business using technology?

Output: 

• Recap of relevant Use Cases
• Strategy Priority List
• Data Strategy Roadmap

During this phase, we will define the design and boundaries of the Data Product. This includes the design of the software and how the services or business are going to be rolled up around the AI and Data product. For example, how the service content, user interface, visual or voice representation of the data product will be. This will enable us to have the overview of the final product before we proceed with the prototype. 

During this phase, we also analyse the first versions of algorithms to give ideas on the possibility and limitations of the products. 

The questions to answer during this phase will be.

1. What does the product do and do not do?
2. How will the business process and team workflow change with the software being developed? 
3. How is the first draft of architecture to navigate further software engineering work?
4. Is it technically feasible to develop and deploy the software in the current legacy system? Is there any integration required? 
5. What are the pros and cons of the current algorithms?
6. What is the definition of requirements to create the software?
7. The estimation of much time and budget required to develop the software. 
8. Are there any other factors that add uncertainty or success  to the project? (e.i. New data sources, new software, new skills required from the team, etc.)
9. What are the potential risks of the project and what are the plans to mitigate them? 

Output: 

• Proof of Concept
• Backlog Definition 
• Software Project Development Plan 

A prototype is a representation of a finished product. With the prototype, we are able to check the experience, flow and content of the software. During the prototyping phase, we also determine the technical possibilities and prepare the set-up such as cloud infrastructure where applicable, databases, repositories as well as accounts. Furthermore, we develop the prototype to showcase the functionality and value of the software. This is the stage where we can validate the ideas as well as the assumptions from the product design phase. In this phase, we answer the following questions.

1. Does the prototype do what it does?
2. Do users think the product’s design matches its purpose?
3. Are the desired technology and user experiences feasible?

Outcome: 

• Product prototype 
  • user experience and content 
  • actual technology on the backend

The test phase reveals the gaps that might have been overlooked during the early stages of analysis. With the end users, we will gather feedback on how the look and feel of the software, combined with the actual technology, provide the desired outcome. The test will give emotional experience about how AI is involved in the final service interaction or business process. 

We will work closely with you to gain feedback and insights on the prototype, while planning the actual development requirements and releases required to complete the software product. 

1. What’s the first thing users would want to do on this product? Can they do that?
2. Is the product self-explanatory? Do the users understand how to use the product?
3. Does the information architecture and navigation make sense? (Can users find what they’re looking for?)
4. Does your target market feel like this product was designed for them?
5. What, if anything, would make your users want to use this product frequently?

Outcome: 

• End user interaction feedback
• Definition of backlog 
• Design of releases

During the development phase, we create a shippable product for you through iterative processes. Depending on the product’s complexity, development occurs in stages, referred to as ‘releases.’ Throughout this phase, we collaborate closely with you to develop application features and simultaneously gather feedback to align with your business objectives. We utilise a combination of technologies to bring your ideas to life, tailored to the desired software product. Explore our solutions page to learn more about our capabilities.

Output:

• Software product (in releases) 

Development and deployment occur simultaneously; however, during the deployment phase, we typically integrate service functions and orchestrate content, data, and functionality for the user. Integration with other systems also takes place at this stage, which may require additional time, especially when obtaining approvals from subject matter experts and ensuring quality and legal compliance according to the business context.

As the deployment phase progresses, the business team begins using the AI software to handle business operations. Consequently, adjustments to the business workflow and processes within the team are made during this phase.

Output:

• Continuous enhancement
• Integration and fine tuning

When development and deployment are completed, the software and related services can go live, marking the beginning of operations. When utilising AI, we typically receive new data, either from application users or external sources. During this phase, we establish a continuous learning and improvement loop or perform manual updates. If you have an internal Data and AI team, we can provide handover training and documentation. Additionally, if you require maintenance support from our end, we offer basic software maintenance support.

Output:

• Maintenance 
• Handover training and documentation

Depending on the complexity of the product and the feedback we gather from the AI product itself, there are always opportunities to upgrade and scale. The insights we collect from the data product are a valuable asset for mapping out opportunities for businesses to thrive using AI.