Where does the future of Enterprise AI lie?
Prebuilt AI engines vs Low code apps
Want to find out more about how prebuilt AI engines stack up against low code apps?
Chat now with the founded.ai team.
Despite economic uncertainty and a reported slowdown in tech funding for some areas, Artificial Intelligence (AI) and Machine Learning (ML) continue to be hot trends. According to The State of AI Report by CBInsights, Q1 of 2022 saw some of the largest M&A deals take place for leading AI companies, and a total of 14 new unicorns also emerged in the field.
McKinsey attributes a lot of this AI growth and interest to Applied AI, where models trained in ML can be used to automate certain activities and make better business decisions. They also point to Quantum technologies, Web3 and next-generation software development as key factors in the future of Enterprise AI, a field they predict to only gather further momentum and interest going forward.
Why is AI of such interest to Enterprise companies?
To understand why AI is taking off in particular for Enterprise companies, you only need to look at where these organisations are investing their time and money, namely customer service, quality control and sales optimisation (according to Gartner).
Across each of these areas, AI really comes into its own, predominantly in the form of robotic process automation, virtual assistants and cloud computing, all of which are brought to life by cloud-based AI and open source development tools.
Furthermore, Gartner point to an increasing adoption of AI because of low and no code applications. According to TechTarget, these are “types of visual software development environments that allow enterprise developers and citizen developers to drag and drop application components, connect them together and create mobile or web apps.”
In fact, low and no code apps are becoming so popular that Gartner predicts some 70% of new apps developed by Enterprises will use low code or no code resources by 2025.
Why are low and no code apps so popular?
Low and no code apps are effectively a great bridge to AI, helping to democratise AI functionality and open up technical fields to people who are often known nowadays as ‘citizen developers’, i.e. individuals without a technical background who want to get more hands-on with the creation of their business apps.
Why is this important? Well, AI is often seen as inaccessible and somewhat of a black box . Add to this the fact that there is a real lack of AI and ML talent available for companies, especially with people who can effectively translate business use cases into AI applications, and you have something of an AI perfect storm.
Although this field is still somewhat in its infancy, startups and large corporations alike are investing in it heavily. Vendors such as Microsoft see it as a great way to tackle resource and talent constraints in the likes of IT and engineering. And plenty of startups are emerging in this market, including DuploCloud, TrueSource, Cyclr and Mendix.
The problem with low and no code apps
Low and no code apps are a great start for non-technical people and departments to understand ‘the art of the possible’. Say the accounting department wants to try out automating a data entry task – apps like Adalo and Kissflow are a great first step to test out that the process will work and how the wider business set-up needs to be adapted around it.
However, when it comes to building out more complex processes, or more tailored solutions that integrate with existing and legacy custom infrastructures, then low and no code apps probably aren’t going to get you very far. For that, you not only need a much wider understanding of how the department functions and operates, but you also need much deeper insight into the technical requirements that will be needed to make any process changes.
Jon Chan, Stack Overflow’s Director of Engineering, puts it nicely:
“One of the seemingly inevitable side effects of making software easier to build is that you tend to sacrifice customizability and a much deeper understanding of how the software works.”
In turn, this can lead to organisational chaos – very much a ‘too many cooks’ scenario!
Prebuilt AI engines - the future of Enterprise computing?
So what’s the solution? Surely there’s a middle ground between complex, custom software development that requires teams of highly-skilled engineers, and low code applications that employees all across the company can access and use.
There is. Prebuilt AI engines.
Prebuilt AI engines are scripts or data models that are packaged up and built according to a specific use case, for example, a recommendation engine for 5 star hotels in a certain region. These engines leverage existing, normally open-source, libraries that are widely available to the public from reputable sources and companies, for example, OpenAI. The libraries are then combined with carefully configured modules and components, and constructed together according to the new use case or business requirement.
Prebuilt AI engines are therefore not ‘off-the-shelf’, ‘build-it-yourself’ models that limit what you can and can’t build. Rather, they are engines that are expertly selected by experienced developers and data scientists who are then able to adapt and customise them to your specific needs.
In other words, the best of both worlds! A faster, more accessible build process afforded by the prebuilt engines themselves, with the ability to customise and tailor more flexibly according to the specific business need.
Like low code apps, prebuilt AI engines open up AI and ML to a much wider audience. However, the effort and time burden isn’t put upon this same audience – it’s kept with the engineering experts who can simultaneously fast-track their development process and get something built more quickly.
It’s an approach that companies such as Microsoft are placing huge bets on – if you want to take a look, here’s their documentation library thus far of prebuilt AI models.
It’s easy to see from this that the future of Enterprise AI and applications really does live with an accessible hybrid model – not as chaotic as low code apps can be, but equally not as unattainable as AI programmes that only the world’s best engineers understand. It’s a path that supports rapid innovation and that leading Enterprise companies all across the globe would be wise to invest in now.
Interested in finding out how prebuilt AI engines can support your Enterprise?
Book a short discovery call with Jonathan, the CEO of founded.ai.