Experience #
Weaviate #
Dates: June 2023 - August 2025
Open Source Vector Database startup / scale-up
SRE Tech Lead #
Started as the first SRE hire, when the whole Weaviate Cloud team was just 4 people.
Finished as the Tech Lead for SRE/Platform (10 engineers).
Key technologies:
- Python
- GCP, AWS, and Azure
- Helm
- Terraform
- Weaviate
- Grafana (Cloud, IRM, and OpenSource)
- Incident.io
- Slack, Jira, Notion, Confluence, Google Workspace
Big projects:
- Commandline tooling (python library + multi-use commandline tool)
- Login, authentication, and kubernetes credential and config management.
- DBA operations (schema, schema repair, admin access, diagnostic endpoints, metrics).
- Backups / restores.
- Disk resizing (for running workloads in k8s).
- Reporting (backups, misconfigurations, resources, fleet inventory, alert history, etc).
- Workarounds for gaps in the internal APIs and backend functionality.
- Homebrew formula for installing bottles from a private github repo (see the relevant post).
- Migrations
- From expensive dedicated infrastructure to shared kubernetes clusters.
- From in-house open source Prometheus + Grafana + Mimir to Grafana Cloud and Alloy.
- Observability
- Created an Observability RFC to clarify observability goals and success criteria, along with a staged delivery plan.
- Deployed a custom (yaml-based) DSL to enable easier definition and maintenance of (Grafana/Prometheus) alerts. This replaced opaque json blobs in unmaintainable terraform configs.
- Built custom CI/CD workflow to manage alerts with meaningful diffs at PR stage and reliable reconciliation at merge/deploy stage.
- Created recording rules to unify diverse timeseries labels into a single consistent schema, enabling a single dashboard to work with any customer workload (previously there were different dashboards for different platforms and regions, preventing any cross-workload comparisons).
- Separated monitoring config from workload deployment, enabling rapid iteration on monitoring changes without affecting running production workloads. This reduced the time to rollout monitoring changes fleet-wide from 2 months to 5 minutes.
- Network Infrastructure
- Enabling gRPC access at scale across existing REST-based fleet.
- Layered Istio Gateway to overcome scaling limitations (on URL Map size) with GKE Gateway.
- Incident Management
- Implemented 24/7 oncall with a volunteer sign-up model using sheets, exported to rotas and escalation config in Grafana IRM.
- Adapted best practice Incident Management processes to suit the needs of the company.
- Created docs and training to help onboard volunteer engineers to the published IM process.
- Team Leadership / Management
- Implemented an introductory form of OKRs (Objectives and Key Results) as a step away from constant churn and firefighting towards an achievable roadmap.
- Hosted regular team retros and gathered feedback on toil / workload.
- Provided daily office-hours for team members and others to be able to get time with me every day.
Gousto #
Dates: May 2021 - June 2023
Lead SRE #
Led a team of 4 SREs. Evolved the operability and observability of Gousto’s AWS technology stack via automation, Datadog, and CloudWatch.
Developed a pythonic client library for PagerDuty to simplify config management, reporting, incident response, and scheduling.
Developed a complete Incident Management approach for Gousto (named IMAG after the same project within Google), created supporting docs and training materials, and delivered training across all engineers in the Technology team.
Tesco #
Dates: April 2019 - April 2021
Principal SRE #
Hired to build an SRE Department for Tesco Technology.
Developed a full-stack SLO reporting system using Splunk, CosmoDB, and Python. Developed and published a full monitoring and alerting strategy for all of Tesco Technology. Implemented an alerting framework for managing Splunk alerts via a simple python-based config language. Defined and shared best practices for managing Kubernetes clusters on Azure via DevOps Pipelines and Terraform.
Google #
Dates: March 2011 - March 2019
Roles:
Senior SRE #
Senior SRE in Mobile Services SRE (Firebase Messaging)
Senior SRE in Network Traffic SRE (GCP Reverse Proxies)
Tech Lead Manager #
Tech Lead / Manager of Apiary SRE (GCP API Gateway service)
SRE #
L4 SRE in Calendar / Tasks / Sites / Workspace SRE
In 8 years at Google I worked with customer-facing apps, internal infrastructure, network management, GCP services, PBs of data, and billions of accounts and mobile clients. Automated Calendar’s 90-step turn up process. Operated fleets of 6,000+ “borg tasks” (aka kubernetes pods). Onboarded Google Tasks. Managed 60+ IPv6 migrations. Moved all custom domain traffic from bespoke proxies to shared infrastructure. Implemented the Server-Timing header in the Google FrontEnd (GFE reverse proxy). Completed over 1000 oncall shifts.
Barclays Capital #
Dates: September 2008 - March 2011
Infrastructure Manager (Quantitative Analytics) #
Hired into the newly created role of Infrastructure and Support Manager for the global Quantitative Analytics group, I was leading a team of three within 2 years.
In addition to daily maintenance, support, and operational activities I was also responsible for managing the weekly release of the QA Library, developing and deploying cross-platform build, test, and release tools, developing a performance testing framework, and providing training within the team and wider group.
Morgan Stanley #
Dates: June 2002 - June 2008
Dynamic Languages Engineer #
Responsible for Perl, Python, Ruby, and R shared infrastructure (including productionizing the dependency management system).
Unix Engineer #
Project lead for several key initiatives in a global team of 20 engineers in London and New York. I started in the SAN / Storage Management area, moving on to own the enterprise host inventory database, DHCP / network management infrastructure, VMWare ESX integration, and the global linux desktop project.
Imperial College #
Computing Department / Systems Programmer #
June 1999 - June 2002, LONDON
Managed all department systems- Databases, DNS, DHCP, Email, Web, NTP, remote access, etc.
Best Applied Paper at LISA 2001 for Lexis Exam Invigilation System.
Education #
Bachelor Degree #
Imperial College / BSc Joint Maths & Computing (JMC)
Dates: October 1996 - June 1999
Result: First class honours. Final Year Project prize.
Interests / Hobbies #
Theatre #
Performing and backstage. Over the years I have been the Acting Director of Imperial College DramSoc; Technical Manager of Network Theatre (Waterloo); a director of the New London Opera Group; “Larry Trotter” in “Larry Trotter and the Bonus of Doom” at the Old Vic (Morgan Stanley staff panto); Production Manager, Stage Manager, Lighting Designer, and Sound Engineer for Bubonic Productions’ sell-out Edinburgh Fringe show: Plague! The Musical
Performer #
Most recently working on my improv skills, I’ve performed with the Barnes Pickles and Slack Captains troupes.
Sound Designer #
2013:
- Calendar Girls
- Harvey (sound design and Elwood)
- Life and Beth
2014:
- Who Shot Santa?
2015:
- Flare Path
- Whale Music
2017:
- Roaring Girl
2019:
- Ringmaster
2025:
- The Anniversary
- A Real Race Around The World
- Under Milk Wood
Stage Management and Production Management #
Director #
Music / Audio #
Drummer #
Composer #
Music for plays (Roaring Girl, Under Milk Wood) and podcasts ( NonFicPod )
My soundcloud: mwyer
Podcasts #
Contributed music and audio engineering for the NonFicPod.
Co-hosted several episodes of the Gousto Tech podcast.
Motorcycling #
I ride a Honda NC750X. Working towards IAM Advanced certification
Scuba Diving and Snorkelling #
Assistant Instructor within BSAC.