It is difficult now to remember what for many people work
was like before personal computers became ubiquitous, but less than twenty
years ago many people went through their working day without having to use one.
Oh yes, dumb terminals had been part of our working lives for many years, the
famed Burroughs TC500, the Nixdorf terminal, and for me, having moved into
international banking, IBM terminals to access payment and the wholesale
banking systems. Having spent some time on admin I'd also been exposed to early
incarnations of Windows, but working on
the queries team in Multicurrency Payments Department (MPD) and sharing
one PC between thirteen of us (used fitfully to log stats and produce the odd
letter), paper still ruled supreme.
 |
| Nixdorf terminal |
 |
| The "green screen" |
MPD comprised three sections[1].
Ref 18, Sterling Payments (payments in GBP within the UK and incoming from
overseas), Ref 24, Currency Disposals (payments in any currency within the UK
and from overseas) and Ref 52, Outward Payments (payments in any currency,
including GBP, going overseas) - which was where I worked. Everything was done
on paper in Ref 52. Payment instructions came in from branches on Telegraphic
Transfer or Mail Transfer forms, payments from other banks arrived by SWIFT [2]or
telex and were printed and re-input manually and all queries involved writing
out vouchers or SWIFT messages which were then entered into the relevant
systems by a separate team. On average we must have had four or five hundred
live files on the go (and thousands that were closed), all represented by
paper, at various stages in the process of being prepared, worked on or waiting
to be filed. Finding one you wanted among these could be a very time consuming
and frustrating process (a phone call from a branch or customer asking for the
current state of play on a particular query could often result in a protracted
search, often fruitless). A solution was needed and so in 1996 (ish, I'm a bit
hazy on dates this far on) The Midas Project was conceived.
Midas (it's not an acronym, just a name) was a project
comprising IT staff and people from the business, some who became involved in
the build and some, like me, who acted as subject matter experts and testers.
Initially I joined the team on a six month secondment; somehow I never really
went back to MPD and at some time during the following six years I ended up as
full time member of the team. It was a
completely different world from the one I was used to. Gone were the daily
deadlines (hourly even) and instead there were deadlines of weeks or even
months to complete certain tasks. It was exciting, rewarding, stressful and
baffling at times and one thing I learned was that you had to be flexible,
especially over working hours. Early mornings, late nights, weekend working and
all nighters came with the territory, especially in the run up to a software
release or during contingency testing.
 |
| Midas Team cartoon, courtesy of Mus Huseyin |
To build a computer system that could replicate the manual
processes in place at the time the first requirement was to define the existing
processes. This meant mapping out somewhere in the region of twenty query types[3]
then designing workflows requiring users answering a series of questions and
generating messages or accounting entries or payments based on the answers.
Generating SWIFT messages to other banks systemically was relatively
straightforward; creating accounting or payments not so. Initially the best we
could do was generate more paper that had to be input into the appropriate
systems; generating entries and payments directly to those systems was
something that came along many years later. Naturally we had to have a query
type for those that didn't readily fit into the usual pigeon holes; we called
it Miscellaneous (original, eh?) and it had no workflow attached, it had to be
worked in what we called "free format." This resulted in a
conversation with a department manager which sort of foresaw the move towards
proceduralising everything and devaluing common sense, expertise and
experience. " How," I was asked, “does the user know how to resolve
these queries?” I replied that they had to use their experience and judgement.
“Oh dear,” came the response, “we can’t have that.” The movement towards
procedures and guidelines outweighing experience and expertise had started.
 |
| I'm middle, right in this one. |
While we now take for granted the PC, the mouse, copy and
paste and toggling between windows, for the average query clerk in MPD in 1997
(the year Midas went live), these were a mystery. Believe it or not, courses
had to be run to teach people about "point and click" and navigating
on screen menus, and for some this came far from naturally. When it came to
teaching the department's staff how to use the system we had built I was in for
a shock as our manager dispensed with the services of the two people originally
tasked with planning and delivering the training and asked me and my colleague
Fakhra Brisby to take it over less than a week before the first session. We
cobbled something together and it seemed to go down quite well, although how
much of it was of any use and how much sank in, I'm not sure.
Midas went live in October 1997 and it is safe to say there
were some issues at first. It was a complete culture shock for everyone;
inevitably most people hated it. Some bits of it worked better than others and
some coding changes and software releases were necessary to address some of the
problems. Probably the best thing you could say about it was that it was no
longer possible for files to go missing. But considering our lack of experience
or a benchmark to measure ourselves by, it went rather better than we actually
thought at the time, so much so that Midas was still going strong when I left
the bank in 2012 (and as far as I know, still is), resisting attempts to
replace it with anything else. The fact that it is now nearly twenty years old
and runs on software ten years older says something creditable about it I
suppose.
Working on Midas was one of the most satisfying things I did
in HSBC, helped to no small degree by working with some great people, like Gavin
Warner, Paul Mills, Steve Tucker, Steve Giles, Gary Cook, Mus Huseyin, the late
Ron D'Castro and Raj Soni, to name but a handful of them (if I carry on this
will be like an Oscar acceptance speech).
If you now spend your working day in front of a computer you
take it for granted, but just think, a little over twenty years ago you would
have been working in an office where copy probably required carbon paper and
paste had something to do with glue.
[1]
Technically I suppose there was a fourth, Ref 62, Drafts, but somehow they were
on the periphery.
[2]
SWIFT - Society for Worldwide Interbank Financial Telecommunication see, www.swift.com
[3]
Typical queries included Beneficiary Claims Non-Receipt, Unable to Apply
Credit, Unable to Apply Debit, Amendment (of payment).
No comments:
Post a Comment