2025-03-27

  • second objective in particular needs work
    • focus on image deployment ; OR
    • self-service

Application image hosting that ‘just works’

Context

Adoption of Cortex has been stubbornly slow. One reason for this is that the platform ‘shifts left’ responsibilities resulting in additional responsibilities and complexity for partner teams. This objective is to offer a solution to take a container image published to artifactory seamlessly into a hosted environment (which might be lambda or K8s), with out of the box answers to things like logging, monitoring, authentication to name a few.

Key metric

Roof:

  • Some blog posts and C3 talks

Moon:

  • End to end application deployment story requiring no more than standard developer tool chains (example: npx/npm for Node developers)
  • Published case study of adoption that onboarded an application in two sprints.

2025-02-05

Support SDLC as it pertains to Ops team

Context

As it currently stands the Cortex Ops team is invested in Python & Poetry ecosystems as it provides the sweet spot between rapid investigation and rigourous software engineering practice.

The wider SDLC process has established maturity metrics across a number of dimensions and the Ops team has set a goal of reaching a secure 3 from a base in the 2-3 range.

Key metric

Roof:

  • Two Poetry plugins created and published that address gaps in the current SDLC support.
  • For example, these might support lambda bootstrapping and deployment or terraforming infrastructure.

Moon:

  • Three+ plugins established as Open Source with at least some initial community established around them. That might be measured by GitHub stars, or code contributions.

Developer advocacy

Context

Adoption of Cortex has been stubbornly slow. Some of this may be attributed to organisational barriers but I believe it is also due to offering more rather than less complexity to the ultimate consumers of the platform: developers.

This objective will focus on simplicity of adoption through providing step-by-step guides that ‘just work’.

Key metric

Roof:

  • Some blog posts and C3 talks

Moon:

  • End to end application deployment story requiring no more than standard developer tool chains (example: npx/npm for Node developers)
  • Published case study of adoption that onboarded an application in two sprints.

Establish Carbon-Ops baseline

Context

Carbon emission reduction is typically driven by three stages: measure, plan and track. In respect to cloud use, there are some quantifications and measurement toolkits available from AWS but as far as I am aware Elsevier is not yet taking advantage of them. Therefore this objective will focus on measurements of cloud use associated with the Cortex platform. It is important to recognise that measurement of Carbon intensity always involves a degree of estimation and extrapolation but strives to be evidence-based and to improve that evidence over time.

Key metric

Roof:

  • Manual estimation of cluster Carbon intensity
  • Ad-hoc findings and recommnedations based on real-world applications

Moon:

  • Automated estimation
  • Business case support for change