That’ll not be too AI for me, please. Cyclr is sailing against the wind, why?
Building on my discussion, Are AI Platforms and Embedded iPaaS a Match Made in Heaven? from about a month ago, I thought I’d double down on where I left off.
Don’t get me wrong, AI is awesome—awesomely powerful, awesome potential. I love it.
At Cyclr we partner with many AI vendors and provide them with a controlled integration ecosystem. This is so they can securely access third-party data their applications need on a customer-by-customer basis. This works very well and Cyclr partnering with AI vendors makes a lot of sense.
However, my decision as CEO is currently not to develop AI inside Cyclr. I recognise this stance is contrary particularly when (and I admire them a lot) other iPaaS vendors such as tray.io have gone all-in and wholesale rebranded as tray.ai.
Reasons for leaving AI out of our Embedded iPaaS
The strongest positive argument for AI within an iPaaS platform is that AI can help with integration flow creation. This can be particularly powerful when users are less technical. As Rich, the CEO of Tray, says in his video, their vision is that ‘anyone can resolve their problems without the constraints of technology’. I am a buyer if it is for your own internal purposes.
So, why not do the same at Cyclr? Well, our audience isn’t users or companies that self-determine their comfort with the use of AI for their own individual/internal data and workflows. Our audience is technology and service companies that use Cyclr to administer workflows. These workflows are pushed out externally to their own customers. In other words, any decision I make to incorporate AI would impact our customers’ customers. We are a B2B2B facilitator, not a direct B2B business.
As of right now, I don’t feel the AI ecosystem is sufficiently evolved to allow AI to participate heavily in a B2B2B ecosystem. I suspect that the use of AI is, and will become increasingly, divisive. Already we have heard rumblings from large corporations that AI must NOT be used inside software products they use and a requirement for full disclosure, and break clauses. If this were to change I can only see this increasing in the short-term. They are not yet comfortable with the ‘black box’ AI segregation and flow of data. Therefore, I think simplicity is best when it comes to large enterprises.
The second reason for my decision is that our audience is technical. We have seen a trend towards wanting more control of the detail and flows – rather than less which AI would provide. This trend even goes against the ‘low-code’ trend of the last few years. At the end of the day, it is my job to listen. Cyclr’s 12-18-month roadmap is full of value-add, more technical, features that don’t need AI in them to be powerful.
What does the future hold?
I fully accept that this is a position ‘of the moment’ for 2024, which may not be the right position for 2026. However, I bet that, a bit like blockchain (that ‘everyone’ was going to use when it came out), AI will find its place in some, but not all, applications.
B2B2B iPaaS (or ‘embedded iPaaS’) where we play is a watching brief at best for now.
