First of all - can it be done?
The commercial use of artificial intelligence offers tremendous potential to businesses of any shape and size. According to PwC, AI is estimated to add $15.7 trillion to the global economy by 2030. This is too much to ignore - whether you are a small business or not. However, many people – from corporate executives to small business owners – have very little, if any, in-house knowledge to build their own AI applications.
We cannot disregard the technology underneath the hood: Proper AI systems require extensive skill not only in data science but particularly in software engineering and cloud computing. But fear not - there is some good news: A few companies have realized that much of this can be neatly moved into a user-friendly frontend for non-technical people.
We want to give you a small taste of what is already possible today and how it can be realized, without the technical noise that often surrounds this matter.
You don't need to know how to code to build an AI model
Most of us are already using AI in some form every day. Be it a search engine, newsfeed, or your phone's Face ID: There is some degree of intelligence in it and the majority of it is certainly artificial. But although we are all getting used to these interactions, creating a system that uses AI is something that still happens on paper (or PowerPoint slides for that matter).
In a similar fashion, websites could be visited long before the general public could create their own websites. Today, any person can build their own website without a single line of code. Companies like Wix, Jimdo, and WordPress provide users with design templates and guidelines. Instead of writing code, people have to shift around boxes, insert pictures, and write text.
These above-mentioned companies have enabled people to leverage the full potential of technology without extensive technical knowledge. Building websites became available to the masses. Local artisans now operate online-shops, the small hairdresser from around the corner offers online bookings.
What works for websites begins to work for artificial intelligence systems as well. Similar to website builders, specialized firms allow people to build their own AI models without code.
What used to be a domain for skilled software developers and hobbyists is starting to become mainstream. And in the same way, as not all websites are built with website builders, we don't claim that all AI systems can be built without code. But we are more than happy to give it a shot!
If you want to get your hands dirty, you can do one of two things: Either directly get started on our platform or dive deeper into the matter first. The following three steps are a good starting point for the latter.
1. Prioritize AI use cases
As with most things, the difficult part is getting started.
We recommend you begin by brainstorming all relevant processes in your company. Write down everything that comes to your mind. Don’t think about complexity yet! To give you structure and inspiration, we built a template that you can use to organize your thoughts:
Now, identify those processes out of your list that are suitable for AI involvement. A suitable process should fulfill the following three criteria:
#1 The task is repetitive → You should use AI to get rid of tedious daily tasks, not to find the new strategic direction for your company.
#2 The task cannot be described by logical rules → If it can, you don’t need AI - go for robotic process automation instead.
#3 A skilled human can decide on the task in a few seconds → AI excels at automating decisions with clear input and related output. If a skilled human needs to think hard about each decision, don't try to automate the task (for now).
If a process fulfills all three criteria it is suitable for AI involvement. In case you want a more detailed overview of how to find processes that AI can automate, we wrote a separate article on that. Have a read here.
Lastly, prioritize your list of suitable use cases. Take into account two factors: importance and complexity. Don’t overburden yourself with something too difficult at the beginning! You should begin with the most valuable option among the processes that are easy to implement.
How to find a suitable AI project to start with.
What follows now is implementation. In a nutshell, your action plan should look something like this:
- Establish a first showcase solution
- Get feedback from your environment (what did work, what didn’t)
- Spark the fascination of relevant stakeholders (e.g. your boss, employees)
- Continue with the next AI project, gradually increasing the process complexity
2. Start a pilot project
Nothing holds you back to start leveraging AI today - neither do you need to acquire any technical knowledge, nor do you need to draft your company's 2030 AI strategy first.
Instead, what you need to do is start building your first pilot project. Implementation certainly is the most important and exhausting part. Even without code, it takes passion and perseverance to give birth to your first AI model.
But it will help you to learn the maximum amount of information in the shortest time. This is what you need, what your boss needs, and, most importantly, what your company needs.
Begin with the highest-ranking use case from your prioritized shortlist. If you have followed the previous step in this article, this is a process that is suitable for AI involvement, has a significant value for your business, and is easy to implement.
Now, look for AI builders that help you to build a custom AI solution without code for that particular use case. With these firms, you don't need technical know-how to put your idea into practice. Whatever your requirements are, there are usually articles or videos that well explain what you have to do.
Our blog, for instance, has a step-to-step guideline on how you can build your first image classifier in no time and without code.
You will soon have your first results. But it is crucial to not stop here.
3. Learn and repeat
It is important to take your findings, learn from them, and then repeat the whole cycle with a new process. Think about the following questions:
- What were your key obstacles? How can you overcome them next time?
- What did key stakeholders (e.g. your boss, your employees) think about the whole process? In your next implementation cycle, highlight the positive and improve on the negative aspects.
- What resources do you need to implement future AI projects more effectively?
When trying to implement a technical project as someone without a technical background, these insights are crucial.
Now, continue with the next project on your prioritized shortlist of use cases. This might be a process a little bit more complex, but also more value-adding.
You will soon find out how far you can get in the world of AI without writing a single line of code.
Let’s summarize the key insights of this article:
- You don’t need to know how to code to build an AI model. Similar to website builders, there are firms offering custom AI models without requiring technical knowledge.
- Start a pilot project today. Don't spend your time learning a programming language or drafting your company's AI strategy - get your hands dirty with a first prototype!
- Learn and repeat as quickly as possible. This is how you gain the maximum information in the shortest amount of time.
- Gradually increase the scope of your AI projects. You will soon find out that you have leveraged AI in your business without writing a single line of code!
So, having finished reading this article, what's next? We have plenty of resources to help you with doing so - But in case you have any questions about this article or we have sparked something in you, feel free to reach out to us via Twitter or LinkedIn. We are happy to help!