AI? Remember Automation…

With the endless AI buzz dominating every technology headline and press release, it’s easy to forget the origins of “computer intelligence”; namely the likes of expert systems, macros and the “A” word – automation.

I remember being involved in many automation test projects in the 90s, often focused around the likes of network management and the origins of “intelligent systems” and networking policies. I even recall writing an “artist formerly known as a column” for Personal Computer Magazine (PCM to its fans) c.’93 talking about how the likes of automation would change the work: leisure balance for many people in the future, so they’d best start looking at extending their gym memberships 😊. That wasn’t me being a latter-day soothsayer with AI in mind – or a fitness fanatic – it was simply what was genuinely going on back then.

At that time, I was doing a lot of work in areas such as file transfer optimisation, sometimes on a general basis (say, ISDN data transfer) and  other times on a specific application (remember Lotus Notes Replication anyone?). Fast forward to 2026 and – guess what – file transfer in every form is still a massively used application – bigger than ever in this cloudy AI world? – and automation is more than alive and well. My latest test project – working with Progress and analysing its brand spanking new Automate MFT product – combined the two; automated, managed file transfer.

For many years now, companies have been smart enough to use managed file transfer (MFT) for their zillions of annual – largely scheduled – file transfer activities, rather than using a manual, ad-hoc methodology (or lack thereof). Progress itself has offered MFT products for many years in the form of its MOVEit technology, but the “cloud clamour” has seen the company combine cloud and automation with the product it asked me to test, the aforementioned Automate MFT.

The full report can be downloaded from here:

https://www.progress.com/resources/papers/automate-mft-report-by-broadband-testing

Briefly, the platform is built around a combination of cloud-based orchestration, agents (cloud or self-hosted) and endpoints (cloud or OnPrem), which interact to provide a potentially global MFT deployment, with scalability on demand – hence the cloud element. Key to the product is the ability to create complex, layered tasks, based on potentially sourcing and delivering data, from and to, several different locations, using different protocols/security types, all in a single, automated job, which can be run standalone or scheduled. And zero programming is involved :😊.

What was an interesting angle to come out of the testing – and one that is an increasingly primary metric in product evaluation – is the TCO (Total Cost of Ownership) aspect, which the report also delves into . For the testing, we looked at a number of real-world use cases, and compared them with the alternative manual scripting approach, with no automation, and produced some very interesting metrics around the ongoing day-to-day costs of doing it manually, and maintaining adds and changes versus the Automate MFT approach of “create once with zero programming skills required and use infinitely with occasional tweaks that take a few seconds”. And that’s before you factor in the cost of “human error”, as well as finding, then keeping hold of trained staff, or having to continually recruit contractors for new coding requirements.

Anyway, you can read all about it via that report link. Meantime, it’s great to know that the “stuff wot I wrote in the 90s” was – astoundingly – correct and more relevant than ever right now. Sadly, my football predictions didn’t follow suit, but you can’t get it right every time!