About me
I'm Graeme Kay. I've spent my career close to the technology: first fixing it, then running it, then building it. Today I'm a ServiceNow developer and administrator, and I use automation and AI to take the routine work away, so engineers can spend their time on the technical part of the job.
How I got here
From walking the floor to building the automation behind it. Each step taught me something I still use.
- Floor walkerWhere it began. I got to know the AV support team because they looked after the hardware in our buildings. I wanted to do what they did, so I set about learning my way onto the team.
- Desk and field support engineerI have loved fixing things since I was young, so a role split between the support desk and the field was a perfect fit. I travelled across Europe repairing audio-visual rooms and installing immersive video systems, from T3 and TX9000 to IX5000, and became the subject matter expert on immersive video endpoints and infrastructure.
- Support desk managerAfter running the desk successfully I was promoted to manage it, and I started moving into infrastructure work.
- Infrastructure leadI took on infrastructure management and became the lead and subject matter expert for Acano, Cisco TMS, Cisco VCS / Expressway, Pexip and Cisco UCM. I also built contingency into my team, which mattered to me before moving on.
- Smart MonitoringWe built a monitoring service that worked the same way across manufacturers, with a custom ServiceNow integration. I started as scrum manager, became development manager, then product manager, and returned to hands-on development for the service’s retirement.
- Moving customers onTo retire Smart Monitoring I built the small applications that moved customers onto a new monitoring service, driven by vendor webhooks into ServiceNow. They now help monitor over 6,000 devices, so customers kept the same excellent service and availability.
- ServiceNow, todayI work as a developer and administrator, with customer service management teams, improving how technology, people and processes fit together.
A closer look: Smart Monitoring
An interval-based monitoring service that watched kit from Cisco, Crestron, Microsoft Teams, Pexip, AMX, Samsung, LG, Atlona and Extron in the same way.
Its custom ServiceNow integration used stateless events, stateful services and links to cases. That let rules apply at different levels to decide what should happen next. In my opinion it was better than a number of the services we see today.
Flapping management
Calm down devices that bounce between up and down, so nobody is paged for noise.
Aggregation
Roll related events up into one major case or problem instead of a flood of tickets.
Outage management
Recognise a real outage and handle it as one, not as many separate faults.
Change management
Know when a change explains an alert, and act accordingly.
Root causes, not guesses
The hard problems are the ones I enjoy most. I work from the logs and a timeline, rule out the probable causes one at a time, and reach for 5 Whys, Occam's razor and Kepner-Tregoe when it gets tricky.
40 of 50 screens shut down
Forty out of fifty screens shutting down.
Cause: A faulty, overheating capacitor.
Echo on domestic screens
Echo occurring on domestic screens.
Cause: Video processing was delaying the audio by more than the echo reference delay, so the echo could no longer be cancelled.
Every seat in the room
I've now been each of the people who touch a service. It gives me a rare view of how technology, people and process fit together, and I work to the idea that everything is a problem. That's not being negative: it means everything has a solution.
Customer contact / user
Internal user
Field agent
Customer service agent
Customer service management
IT agent
Developer / administrator
My goal now is to automate processes and administration intelligently, with as little interaction from engineers as possible.
Want to work together?
Tell me what you're trying to do and I'll tell you whether I can help.
Get in touch