Bergen, Norway

2024 Conference
Jaeger
Prometheus
Kubernetes
Falco
Harbor
Cilium
Argo
Vitess
Helm
Argo
Cert Manager
Falco
Kubernetes
Cert Manager
Shipwright
Kured
Crossplane
Containerd
Operator Framework
Vitess
Backstage
Linkerd
Argo
Etcd
Kubernetes
Virtual Kubelet
Harbor
Kubernetes
CoreDNS
gRPC
Flux
Etcd
Crossplane
Helm
CoreDNS
Flux
Cilium
wasmCloud
Cert Manager
Cert Manager
Kubernetes
Istio
Istio
Envoy
Shipwright
Kyverno
Etcd
OpenTelemetry
KubeVirt
Latest from Bluesky

Noah Hall

Tech Enabler @ Schibsted | Board Leader @ Tekna's Network for Developers

About

In my day job, I'm a "tech enabler" in the CTO Office of Schibsted Media.

In my volunteer job, I'm the board leader of Tekna's Network for Developers.

For fun, I'm a programming language author (Derw, NeuroLingo, Gwe, Stalk, Scramata).

And I have 3 active blogs where I write about: the tech industry, staff engineering, and news media.

Session

Running a company-wide post-mortem

Presentation (40 min)BeginnerEnglish

Session Recordings

Session Recording

Schibsted Media is Norway and Sweden's largest media house. Composed of multiple media houses (Aftenposten, VG, BT, PodMe, Aftonbladet, Omni, SvD), our technical architecture has gone through some rough periods. Do we centralized everything? Do we provide software as self-service? Are our newspapers actually different enough to need new teams? In 2016, the goal to centralize all our different stacks between the media brands begun. In 2023, we realized it wasn't working.

I was brought in to perform a post-mortem with the entire organisation to figure out the good, the bad, and the ugly. After hours of retro sessions and interviews, we wrote a report that's 1/3rd of the length of the first Harry Potter book. This document has become our most read technical artifact - shaping our most recent re-org and team decisions.

In this talk, I'll share with you:
- The process we took to understand what went wrong
- Some of our findings
- And how we've addressed those problems

Expect to hear no corporate speak: just honest voices on what happened. If we can't learn from the past, we're doomed to repeat the same mistakes.

You should come away knowing:
- How to conduct a post-mortem at scale (400+ employees affected)
- Understanding how to use empathy and listen to voices
- Know how to package those voices and learnings so that management can use them / act on them
- The technical problems that we uncovered from our work

Architecture & Software Patterns