The Narrative

An imperfect guide to building the ultimate digital research assistant, the Under Cloud: Part 1

Yes, this is a bold claim, but it helps to aim high and to cultivate some self belief in what we do. So if it’s not the ultimate now, the goal is to — paraphrase the great Captain Jean-Luc Picard — make it so!

Building the ultimate digital research assistant, the Under Cloud, like anything else — relationships, confidence, a house of cards — requires commitment, perseverance, self belief, among other notable qualities.

Building things also requires time, too — lots of time. In truth, I would have known these things at the time, but it’s the path I’ve in part followed and to some extent been forced down that’s been the most difficult part of this voyage.

“The ultimate digital research assistant?” You asked.

Yes, this is a bold claim, but it helps to aim high and to cultivate some self belief in what we do. So if it’s not the ultimate now, the goal is to — paraphrase the great Captain Jean-Luc Picard — make it so!

An end to the beginning

During the summer of 1999, I’d received an object lesson in how not to manage people, projects, along with most other things related to running a business.

I had, however, anticipated that change was due, so in June I incorporated Octane Interactive Limited, which would be a vehicle for that change, leading to freedom, financial independence et cetera et cetera.

In spite of what became a liberating new beginning, it had be born from an acrimonious end, one I’d rather not have experienced.

While studying for a degree, I’d been head hunted, picked out of class, and within a week working in Leeds, an enviable position for most aspiring students, or anyone hoping to level up their game.

For almost 4 years, I worked from offices in an old building of considerable character, once a mill, nestling between a pub overlooking the River Aire, and an independent department store, not far from the historic and bucolic Kirkstall Abbey. I made good friends, had some great times, learned new and valuable skills — but like most good things, it wouldn’t last.

I was 24, and I had zero intention of working for anyone again.


2002 arrived, and I found myself at a crossroads. While working on the website for a local college, the task of building the umpteen pages for their curriculum was both lucrative and a stunning bore — something had to be done, but what? So, after a bit of research, PHP and ASP seemed to be the two most viable, practical choice of programming language that would allow me to write dynamic web pages, to access data stored in a database.

In the end it was a simple choice, and PHP became the future programming language of Octane. However, the experience of doing that basic research was less than optimal, and I couldn’t shake the feeling that something was missing or wrong.

Context is everything, and it’s worth keeping in mind that the world wide web was nascent, and — in places — it was a fantasia of iframes, invisible GIF files, 8 bit graphics, Internet connections that were so slow semaphore looked like a viable alternative. For reference, we would use tables for certain types of layout because CSS wasn’t up to the task.

But the point is, this was the equivalent of the Cambrian Period of the web, where from almost nothing almost everything emerged, and the humble link became omnipotent, from which would spring a whole maddening and frenetic ecosystem of (in)sensible technologies, most to be abandoned, laughed at and then abandoned, or — of those few that remained — adopted and allowed to mature into the venerable services and software packages we use in this modern era of the web.

As the business progressed and the work became more coherent, structured, and purposeful, I kept bumping into the same damn problem of not having some sensible workflow for doing research. I was writing notes, finding relevant web pages, but I needed to link these things together, which — to me, at least — seemed to be so obvious, but I was damned if I could find anything that would do that one simple thing.

I had heard rumours that IBM had some internal behemoth (think Microsoft SharePoint, but for engineers wearing pocket protectors and sock suspenders) that did something similar to the Under Cloud, but that remained a rumour.

Valley of Jobs

It was in and around 2004 that I’d been working on another idea. So think Basecamp, but it would have been more of an open source environment for people to build their own applications. Although I didn’t know for certain at the time, I was beginning to suspect the ideas I had weren’t so much limited by my own imagination or abilities (growing, but not kung fu), but the scant resources I had available to me, and the environment I found myself in.

This wasn’t Silicon Valley, and it was not the Thames Valley, either, it was the Dearne Valley, and the future was still a twinkling artefact on the distant horizon.

In fairness, the region was once a thriving if blackened landscape of coal mines and steel production, that had long since been swept clean, and along with it the notion of a job for life. Instead of the hustle and bustle of pits and the endless ebb and flow of trucks and men, a verdant landscape of fast-growing trees and shrubs prevailed, remediating the harsh shale of the slag heaps beneath them.

Large-scale employment was the focus of the business development agencies, and not so much the fanciful thinking from the likes of me.

60 months and counting

Between 2002 and 2006, Octane was “fire fighting” for businesses in the Yorkshire region. You see, cowboys don’t wear stetsons in Yorkshire — at least, they didn’t up to 2006, and instead had fliphones, used DreamWeaver, and played buzzword bingo — and the cowboys in this case would promise the world to naive businesses hoping for a website that would deliver unto them the masses, but leave them thousands of pounds out of pocket with a lot of unanswered questions.

I’ve sort of slipped ahead of things a bit here, because it was 2004-5 that I’d given some serious thought to the research problem. I’d tried building something from the menagerie of bookmarking services of the period ( was the main contender at the time), tethered to some other service (Yahoo! Pipes, the IFTTT of the time) that funnelled data of one type into something else that allowed me to keep notes, but this whole Heath Robinson / Rube Goldberg contraption was an ungovernable mess. I needed to dig deep and write an application for myself.

As luck would have it, I’d bagged a promising new client, and I had the chance to develop a framework onto and around which I could build this newfangled research wrangling thing. So, I talked with the client and told them that I’d forego the development fee for the R&D phase of the project if allowed to use that framework as the basis for both their application and the one I was working on, and they agreed.

The prototype was such a colossal, laughable, and chaotic joke that I sulked for an entire 18 months. It wasn’t until a friend, Anna, stepped in and gave me a bit of nudge that I roused myself from the majestic state of self-inflicted ennui, and I began with a complete rebuild.

A point worth noting is that Octane was at this time in its fifth year of incorporation, bucking a significant trend, one that predicted most businesses wouldn’t make it to their first three.


2008 was a time of fitful but often furious development, where professional foci would vacillate between working on the Under Cloud and client projects, and since I was gaining genuine use from it, it seemed important to keep making improvements to it: bug fixes; new features; testing; and using the entire process to help hone skills that had immediate transferrable value to Octane’s regular client projects.

It was during this time that I came to a realisation — the 20-80 rule, where 20% of the clients I had were responsible for 80% of the noise. On the face of it, ejecting the 20% who were dead weight made perfect sense, but what didn’t make perfect sense was the timing, and I almost went out of business — I didn’t hit the financial crisis of 2008 so much as it hit me and the remaining 80% of Octane’s clients so hard that some shrank while others blinked out of existence. Yet, I somehow survived while others didn’t.

2014 arrived and I’d been using the Under Cloud for some time. I’d developed it for selfish reasons, and — at the time — it did 70-80% of what I needed it to do.

“Where did the name ‘Under Cloud’ come from?” You ask. Ah, the name.

The thing is, the whole cloud thing was a new thing (to me, at least), and the web thing had become a passé thing, but that made no sense to me, since most of what was useful was still sat there in the web. So the idea for the name was a sort of tongue-in-cheek homage / reference to the enduring usefulness of the web in spite of the mass cloud grab that was going on.

Anna, who gave me the important nudge was studying physics somewhere in Canada, but was also tracing her family tree, and how — to much amazement — used the Under Cloud to create the genealogical map: first with folders to group parents, and notes for each parent, linking each to web pages containing historical information; and then using links between the notes and web pages representing a person to trace other notable connections.

At this point, I was using the Under Cloud week in and week out, making improvements, fixing bugs (while creating new ones), and — to use rugby parlance — moving the gain line forward with each iteration.

There I was, in one meeting or event after another and I’d have one client or delegate after another peering over this shoulder or that, looking at what I was doing.

“What’s that?” They’d ask.

“Oh,” feigning nonchalance, “it’s — eh — something I’m working on. Nothing much.” I’d confess with an inward chuckle as I’d link a note to a web page with a suitable contextual description.

After a while and several ad hoc demonstrations, sufficient praise, gasps of “Wow, that’s good!” I knew that I had to up the Under Cloud’s game, and get it fit for prime time…

So, the planning again. I am impatient; when I set myself a plan, I want it done then, not tomorrow, next week, or the month afterwards. I decided to approach RISE:

RISE is a unique business support initiative delivered in partnership by Sheffield City Council, Sheffield Hallam University and the University of Sheffield, and is part-funded by the England European Regional Development Fund as part of the European Structural and Investment Funds Growth Programme 2014-2020.

… and — as usual — I missed the cohort by milliseconds, or something daft like that. However, as luck would have it, I managed to snag the interest of couple of excellent developers, and since I’d not done the recruiting thing before, I took the pair of them.

Within a couple of weeks, the Under Cloud was in for a major overhaul!

It’s a shame — and although I knew it at the time, given their prodigious talent — that I couldn’t hold onto either of them, Dario and Sina, for as long as I’d hoped, but the progress made was sufficient to push the Under Cloud far down the track, giving me the head start I needed to plan for the next pit stop.

A pivotal moment

2016 came, and after a some inward examination, I knew that I had to do something different — something that would scare me.

Now, I’ll be the first to admit that getting an office, when placed on the spectrum of scariness, is some distance from climbing the north face of the Eiger, or waging war in the Middle East, but for me it was a substantial change in personal, professional, and financial circumstances.

Working alone isn’t a problem in and of itself for someone like me, I’m more than capable of figuring things out and doing everything for myself — I’d made it a point of differentiation to compete with businesses comprised of teams, and out perform them, accomplishing more in less time. I’ve had the type of upbringing that has forced resilience and resourcefulness onto me.

Working alone for long periods with little contact with anyone is a weird and curious thing, one where I sort of descend inward, talking to myself, forgetting the simple etiquette of human interaction, the ebb and flow of conversation, challenges and objections to ideas, and the nearness of someone else.

But over time, a sense of incompleteness builds, like the awareness of thread thin cracks in a much loved vase, or the mottling at the edges of an old mirror, things not visible at a distance.

Luck intervened as a force of good (for once), and the shared space happened to contain two people, Chris and Adam, who I would go on to consider good friends, and Adam in particular would have a transformative effect on the direction of the Under Cloud, insofar as turning the path ahead from something akin to a shotgun blast in a China shop into a precision takedown of an infant tardigrade with a Russian T-5000 Tochnost rifle from 1,000 meters.

Because I’d been using the Under Cloud for personal and professional reasons, it had a perfect market fit — as in, I was the ideal customer. But the plan was to go beyond me, and that would require a new audience. I needed to pivot…

In part 2, the Under Cloud & I undergo a series radical transformations.

Logo for the Under Cloud, the missing link in the evolution of doing research.
Under Cloud is the missing link in the evolution of doing research.

Our brains aren’t hard drives, and how we remember things and make connections between them is personal.

Under Cloud is the missing link in the evolution of doing research — learn more.