Perfect timing to start your journey into AI.
While large corporations benefit from the enormous amount of data available to them, having access to expertise in business, technology and regulation, most SMEs dont. The results are often denial: To simply label AI as a big hype that will pass over and make a place for the next big thing. The amount of solutions already implemented and fully integrated into your life, suggests denial is no option.
For those SMEs who want to explore what AI or machine learning can do for them, to either transform digitally, automate production, develop new services or create value from their data, these are the resources you need to develop and posess: Domain knowledge (read: People), technology, data and processes.
AI competence, culture, processes, skills, and mindset must be developed internally in the company to get real future value. If you need outsource, or use consultants, collaborate tightly, be actively involved in the project and use the external competence as a good advisor that helps you see the opportunities that take you toward your goals, while developing your own competence.
To get started with an AI project there are several considerations to be made around business and leadership. Machine learning and digital transformation must be defined as an input factor in the company's business strategy with some clear goals on what to achieve! Do you need to defend your market position, do you need to catch up with the competition, – or do you se a possibility for disruption? These questions must be answered before any sort of AI project is set in motion. A pilot case to verify data quality and explore value options to guide strategy work is a good way to start.
Selecting the right case is crucial. Sales, Marketing, Accounting, HR or production? What question will give the right insight? Don’t start with an easy one just because you think it is an easy starting point - you need to extract real value out of the project and estimate ROI – the job and the AI journey will be the same anyway. Start small, pilot problems, fail forward. Businesses can realize immediate incremental value with narrow pilot efforts that are focused on solving concrete problems increase value by learning from mistakes. Starting small doesn’t mean thinking small, instead, every AI initiative must begin by assessing data, people, and processes to support and align with broader business objectives, impacts and governance structures. Therefore, you must identify and prioritize ideas to work with where you have historical and available data. A specially adapted design thinking workshop would help you get there.
In our design thinking workshop, you work with experts in design, architecture, agile development, data science and AI, automation, blockchain, security and business development. In the modeling phase you will develop an MVP as a proof-of-concept, to arrive at a scaled-down sketch for the core of the solution: the machine learning model. During this process you will identify business opportunities and models, explore possible paths of value creation and surplus value, and end up with an estimated ROI and a clear goal for the project.
Machine learning and AI allow you to make better use of your unique knowledge of your business and market, supported by your available data. A typical project includes four key areas to consider regarding data: how to collect, organize and analyze data, and ultimately apply AI - data maturity. (The AI Ladder).
Once you have picked a question to answer, your domain knowledge will tell you what types of data might help answer your question. Based on your knowledge of your process, you'll know that some types of data are very important, and some types are not, and that is going to be very crucial for getting good answers out of your AI project. Therefore, a defined project plan with a clear budget is difficult to put together – why? Mapping and inspecting the available data and identify obvious problems in the data, such as gaps, noise, lack of structure, little data, and other challenges will affect time and effort. Many such challenges can be solved, but this can be time consuming work. Note that available historical data is "make or break" for an AI project.
The next and last phase is to train the model and when it is become good enough to perform the desired task at an acceptable level, it can be operationalized in the organization.
And remember that an AI project cannot be compared to an ordinary IT project or process, it requires a different approach, both in planning, execution, and budgeting - the journey is the desination. Only when the trained AI model is on place, made available to users or other IT systems in a scalable infrastructure, secure and easy to operate, it becomes a traditional IT project.
The perfect timing is now:
If you are a Norwegian company – you should also explore your organization’s possibilities for funding from Innovation Norway (IN) to get started on your AI journey. A preliminary study must often be carried out in advance of an innovation project, where the purpose is to clarify important preconditions for the project. Generally, IN expect the company to finance this phase itself. In connection with the ongoing corona crisis, they allow companies to apply for funding to conduct feasibility studies. The maximum grant from IN is NOK 500,000.
The purpose of a feasibility study is to provide answers to significant questions related to technical, organizational and market conditions, before any main project – which can also receive financial innovation support. The results of the feasibility study provides a basis for decisions on implementation of an innovation project – and AI is innovation!
Thanx to Karl Ove Aarbu and Trond Kathenes for their inspiration.
