In late 2008 I was working as a freelance Python developer. I’d recently started to use Twitter and among the various “famous” people I had followed was Tim O’Reilly. One of his tweets caught my attention; it was about an interview that tech-blogger-about-town Robert Scoble had recorded with a hacker who was building “something amazing” called Fluidinfo (so amazing that Scoble described it as “the unfundable world-changing start-up”).
Here’s the summary notes I jotted down while trying to make sense of what I was watching:
- A re-imagining of a data store in the cloud.
- Openly writeable (anyone can add data a la Wikipedia philosophy).
- Simple to find useful information by mashing up data from many different contributors.
- An evolutionary model of data / emergent schema.
- Bottom up rather than top down.
After a couple of hours of unlit, uncut and rather shonky video shot in the hacker’s apartment in Barcelona, my first thought was, “this hacker’s either a complete nut-job or he’s on to something”.
I had to find out more.
It turned out that the hacker was called Terry Jones and, as they would say in the Matrix, went under the “hacker alias” of terrycojones.
Terry, it appeared, had balls.
A quick look at the biography on his website indicated that he’s actually Terry Jones BSc MSc PhD ~ unicyclist (see below), mathematician, chess player, computer scientist, virologist, writer and entrepreneur (not necessarily in that order). I sent him a long rambling email asking for more information about the [r]evolutionary nature of Fluidinfo.
Three months later I got a reply with detailed answers. There was obviously substance to the ideas and claims first made in the Scoble video. At the end of the email Terry invited me to proof-read the developer documentation for the soon-to-be released Alpha version of Fluidinfo. Contained therein was a thing of beauty: a simple RESTful API backed up by an elegant philosophy of how a data store should behave.
I was hooked.
I was also going to the next EuroPython conference in 2009 and so, it appeared, was Terry: the very first talk of the very first session was called “Introducing Fluidinfo”.
Terry and I chatted for a while after his presentation and I learned that he’d been thinking about and planning Fluidinfo for ten years. It quickly became apparent that Fluidinfo was a labour of love. Terry had sold his apartment to finance development of Fluidinfo, given up a career as a computational biologist at one of the world’s leading universities to work full time on it and was in the process of getting a round of funding sorted out before the money (and any chance of getting his apartment back) ran out.
Terry deserved his “hacker alias”.
As with many developer conferences, some of the best stuff at EuroPython happened in the “corridor track”, that setting halfway between Josette’s O’Reilly book stall and the coffee bar. It was serendipity that had me sit down in the corridor next to someone I initially thought was an Irish hacker (by the look of him at least: he had a ruddy complexion, a reddish beard and was wearing green – to me that looks “Irish”, perhaps I was thinking leprechaun). I was wrong, the guy was from Barcelona, a member of the Apache Software Foundation and called Esteve. We had a long and often funny conversation at the end of which I asked the inevitable, “so who do you work for?”. His immediate answer, delivered in the same way that Inigo Montoya introduces himself in the Princess Bride, was “I work for Fluidinfo and we’re going to change the world”.
Terry wasn’t the only one connected with Fluidinfo who had balls.
It turned out that Robert Scoble was wrong: a few months later Fluidinfo received funding through Betaworks and can count Joshua Schachter (Delicious founder), Esther Dyson (journalist and investor) and Tim O’Reilly (whose tweet originally brought my attention to Fluidinfo) as investors. Last year, when Fluidinfo won the “Best Technology” category at the Launch Conference, Scoble was heard to say, “I was wrong”.
Having high profile investors is very helpful. For example, during an interview at the South by South West gathering Tim O’Reilly named Fluidinfo as his favourite start-up. Such PR gifts are usually followed by a flurry of blog-posts and tweets from people who, upon discovering Fluidinfo, announce how amazing the concept is. A typical example being, “this should blow your mind: Meet Fluidinfo, the most disruptive start-up you haven’t heard of”.
Herein is perhaps the biggest challenge facing us: despite being lauded by the tech-cognoscenti we need to take the abstract philosophy and existing technology that underpins Fluidinfo and make it available and understandable to everyone.
To this end we’re working with and learning from various large companies who would like their own instance of Fluidinfo running within their organisation for the purposes of data sharing and information management. We’re also listening carefully to our existing “early-adopter” users and introducing new features where required. Furthermore, we’re experimenting with a web-based front-end for non-technical “civilian” users. We hope that the fruits of our labours will be arriving within weeks.
In the meantime, there is an O’Reilly book called “Getting Started with Fluidinfo” (with a magnificent Flying-Spaghetti-Monster like jellyfish on the cover) which explains the API and philosophy behind what we’re building. Alternatively, a simple technical introduction to Fluidinfo will be appearing on this site very soon.
(About the author: Nicholas H.Tollervey works for Fluidinfo and co-authored the O’Reilly Fluidinfo book. He’s a classically trained musician, philosophy graduate, teacher, writer and software developer. He’s just like this biography: concise, honest and full of useful information.)