CHRONOLOGY
Timeline
Not every job. Not every project. The moments that show how the interests evolved, and how each wave of technology changed the questions.
Autonomous software, bounded authority
DaemonCore and agentic software engineering: identity, authority, operating context, evidence, governance and how to let software act without quietly handing it the keys to everything.
The web, but for agents
Semantic Publishing Protocol and the AI Starvation Loop shifted the publishing question from “what does the page look like?” to “what meaning, attribution, intent and consent should survive when software consumes it?”
Product, architecture and longer-form teaching
Attollo product and delivery work continued alongside workshops and solution-building sessions. Nielsen Norman Group listed the Attollo team among its 2021 Intranet Design Annual winners for Keysight.
Remote work becomes everybody’s problem
The Isolation Conversation brought together people from Microsoft, Deloitte, Avanade and the community to discuss remote work, wellbeing, education, development and what came after lockdown.
PowerUp365, Teams and digital workplace
Filtering the Microsoft 365 update firehose, exploring Teams as an application surface, and building maintainable Power Platform solutions rather than disconnected apps.
Automation patterns, not just automation demos
Microsoft Flow Conference, PnP community calls and a run of technical writing on error handling, expressions, integration and architecture. The theme increasingly became “how does this survive production?”
Power Platform, with a detour into VR
PowerApps and Flow became a serious focus. In October I delivered a developer session from AltspaceVR and built the app through a browser inside the headset. A terrible user experience; an excellent experiment.
Microsoft MVP & Red Plane
Microsoft MVP recognition, launch of Red Plane, and a run of public cloud-adoption material. The Geneva Office 365 session reached 109 slides because apparently brevity had not yet been invented.
Community leadership and cloud transition
Leading the North West SharePoint User Group UK community while Office 365 was moving from curiosity to serious enterprise decision. The talks shifted from product features toward trust, privacy, control, architecture and migration.
Large-scale SharePoint work
Enterprise collaboration, search, information architecture and migrations. Public decks from this period include SharePoint 2013 Search, first-time adoption, solution design and the story of a large-scale upgrade.
SharePoint, documentation and the start of a public career
I returned to contracting because travelling had emptied the bank account, not because I had decided to recommit to technology. SharePoint 2007 changed that. It was a major leap from SharePoint 2003 and good documentation was scarce, so I joined the London SharePoint User Group, started blogging what I learned and found myself pulled back in by the unanswered questions.
After moving to Manchester I started the North West group, then began speaking, bringing other specialists into the community, mentoring new regional organisers and building the infrastructure that connected the groups. That work eventually led to national SUGUK leadership and Microsoft MVP recognition.
A deliberate reset
After years as a developer and contractor in London, frustration with budgets, deadlines and the compromises of project delivery had turned into disillusionment with the industry. At 28 I stepped away: a snowboarding season and instructor training in Whistler, followed by solo travel through Southeast Asia and Australia.
I came home wanting a life built around snowboarding in winter and Divemaster work in summer. Technology was supposed to be a temporary way to rebuild the money I had spent travelling.
Contracting, London and the limits of project delivery
After Precision Web closed I moved into contracting and then London. Years of development work taught me how much software quality is shaped by forces outside the code: project budgets, deadlines, scope and organisational compromise. Eventually the gap between the quality I wanted to build and the quality projects allowed became part of the reason I walked away for a while.
Precision Web: founder lesson number one
During my final year at Reading, I was already doing paid project work for Microsoft UK with Bez Baker, an experienced permanent Microsoft employee, and my fellow student Kevin Bluer. We built internal software including UKWeb, Microsoft UK’s intranet, Exchange Digital Dashboards and work using Microsoft CMS Server. I also spent time teaching Flash and ActionScript to Microsoft staff.
On graduation I turned down the BT graduate route I had originally wanted and co-founded Precision Web with Bez and Kevin. For a 21-year-old graduate it meant building real systems and interacting directly with senior people inside Microsoft almost immediately.
Microsoft was also effectively our only significant customer. When the work was internalised, the business no longer had a viable base. Precision Web closed cleanly and everyone was paid. It was an early lesson in customer concentration, and in recognising when a business should stop rather than assuming something will arrive to rescue it.
Applied Computer Science, cybernetics and the questions underneath AI
At the University of Reading I gravitated toward the Cybernetics department and the work around AI, robotics and machine understanding. Professor Kevin Warwick was a major influence. One lecture question stuck for decades: “What is a chair?” It was a deceptively simple way into classification, meaning and what it actually means for a machine to understand something.
My original final-year project was an ELIZA-style conversational program written from scratch in Pascal. I read into natural-language processing and tokenisation, then made the useful engineering decision that it was too ambitious to finish properly in the time available. The replacement project was a sensor-driven home automation system with a robot arm.
Integration testing and Y2K rollback
My year in industry at BT put me into Directory Services integration testing. I also wrote the rollback guides for Directory Services Y2K incidents. It was an early lesson in a question that would keep returning throughout my career: what happens when interconnected systems fail, and how do you get safely back to a known state?
Computing, encryption and software people actually used
I was not allowed to take Computing at GCSE because I already had too many technical subjects, so I asked for the course material and completed the reading and exercises anyway. At A-Level I finally got to take Computing formally, studied subjects including PGP encryption, occasionally found myself explaining parts back to the teacher, and finished with an A.
Before university I also built two Microsoft Access applications for my dad. One registered and tracked vehicle number plates at the prison where he was Head of Security. It was not a school exercise: staff used it operationally for years.
Taking the black boxes apart
Electronics kits, switches, resistors, lights and building an FM radio on a children’s breadboard came before the professional technology. So did taking broken electrical things apart, trying to understand the VCR, typing BASIC listings from Commodore 64 magazines and waiting for the next issue to publish the corrections.
One of my first original programs, probably on the Amiga, predicted the colour of budgie offspring from the two parent birds. The implementation was a spectacular pile of IF statements. It worked.