You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
2026-07-02 Green Software Principles & Patterns agenda #415
changed the title [-][DRAFT] 2026-07-02 Green Software Principles & Patterns agenda[/-][+]2026-07-02 Green Software Principles & Patterns agenda[/+]on Jun 30, 2026
The Patterns Project held its regular working session on 2 July 2026, bringing together the project lead, operations leadership, and members from contributing organisations including AVEVA and Goldman Sachs. The team reviewed execution of the Patterns 2.0 initiative, validated the effectiveness of pattern proposal review processes, and advanced two critical pull requests through the approval pipeline. Naveen's nine AI patterns are progressing toward merge following incorporation of Green AI Committee feedback. Pratik's new Edge AI pattern has been routed for technical review by domain experts. The team also discussed strategic next steps for catalogue harmonisation, welcomed a new AVEVA contributor, and confirmed continued effectiveness of the pattern skill for proposal refinement.
Pratik from AVEVA submitted a pattern on optimizing power performance for Edge AI inference and training. The pattern was run through the pattern skill, which generated targeted feedback on scope, focus, and coverage areas. Pratik incorporated all feedback, and the pattern is now ready for formal approval. However, Liya raised an important quality gate: AI-domain patterns should be reviewed by the Green AI Committee to validate technical accuracy and domain expertise, not just structural compliance. Russ and Liya agreed to engage Goldman Sachs colleagues on the committee to arrange this review. Once technical sign-off is received, Liya will approve and merge. This review can be raised at the Green AI Committee's scheduled call next week.
Naveen has developed nine AI patterns as part of the foundational work on AI sustainability. The patterns received detailed feedback from Christian at Nuremberg University (Green AI Committee member), which Naveen has incorporated. Russ and Naveen discussed whether to seek further feedback before merge or proceed with an iterative approach. They agreed that Christian's feedback was comprehensive and substantive enough to proceed to merge, with the understanding that further iterations can happen post-merge if needed. Russ will review and close out all comments in the PR; once complete, Liya is ready to merge. Paula raised a design consideration: the nine patterns span multiple lifecycle phases (Architecture through Development), and merging a large batch at once could create category imbalance. The team acknowledged the risk and agreed to monitor for any need to create subcategories (e.g., splitting AI patterns by inference vs. training).
Pattern Submissions & Skill Effectiveness
Beyond the two PRs, the team reviewed incoming pattern proposals from multiple sources, including a new collection of green arrow patterns and additional AI-related patterns from AVEVA contributors. The pattern skill continues to demonstrate value: it catches common issues such as overly broad scope spanning multiple domains, missing sections (particularly cost impact), and insufficient grounding in user requirements. The team confirmed this validation process is working well and will remain standard for all future submissions.
Reference Implementations & Embedding Patterns in Tooling
Naveen Balani shared a green code skill he has developed as part of the foundation's reference implementation work. This represents a key strategic direction: bringing patterns to life by embedding them into tools and workflows that developers actually use, rather than leaving them as standalone documentation. The team recognised this as important for adoption and discussed how to scale this approach across other pattern domains.
Catalogue Recategorisation & New Lifecycle Framework
The project has completed a major recategorisation of the patterns catalogue against a SWEBOK-based lifecycle framework (Requirements, Architecture, Design, Development, Testing, Operations, Decommissioning, End User). This restructuring has clarified where patterns fit within a complete software lifecycle and created more intentional homes for contributions. As a result, several patterns proposed in 2023 — which previously had no obvious category — now have clear placement. For example, architecture patterns were requested in 2023; the new Architecture category now provides that space. The team will revisit these older proposals to validate whether submitters still wish to contribute under the revised structure.
Catalogue Cleanup & Consolidation
Russ Trow has completed an assessment of proposed catalogue changes (deletions, renames, and consolidations) with detailed evaluation of validity and implementation approach. These recommendations will be reviewed offline by the project lead for approval, then prioritised at the next meeting. Alongside this, the team will triage the backlog of previously proposed but unresolved patterns to determine which should be advanced, which should be consolidated, and which are no longer relevant.
Design-Phase Tooling & Accessibility
The team discussed whether design-centric tools — specifically Claude Design and related AI-assisted design platforms — could support the creation and refinement of design-phase patterns. This opens the possibility of making pattern contribution more accessible to non-engineering roles (designers, product managers, UX practitioners) who may not be comfortable with current proposal workflows. The team noted that the design tools landscape is evolving rapidly and took an intentionally agnostic approach: rather than lock patterns into a specific tool, the goal is to develop a portable methodology or wrapper that contributors can apply using whatever tooling they prefer. This avoids vendor lock-in and maximises accessibility.
Contributor Engagement & Growth
A new contributor from AVEVA joined the project for the first time, bringing architectural expertise. The team welcomed broader engagement from member organisations, with several cross-organisational connections already forming (e.g., AVEVA contributors active on both the Patterns project and the SCI for Web initiative; GSF team members meeting AVEVA sustainability leads at external events like London Climate Action Week). The active submission of patterns from AVEVA (Pratik's Edge AI pattern, Babu's new participation, and broader engagement) suggests successful member recruitment.
Key Decisions
PR Optimize power consumption for edge AI inference #414 will be routed to the Green AI Committee for technical review before merge; Liya and Russ will coordinate with Goldman Sachs committee members to arrange this.
Pattern categorisation for the nine AI patterns will be monitored to avoid category imbalance; subcategorisation may be required if any lifecycle phase becomes overloaded.
The pattern skill will remain in active use for all future pattern proposals, validated against the standard review criteria.
Catalogue cleanup assessment will be reviewed offline by the project lead, with prioritisation decisions made at the next meeting.
Previously proposed patterns will be systematically triaged against the new SWEBOK lifecycle framework; submitters will be re-engaged where current status is unclear.
The team will continue to explore design-centric tooling options, prioritising an agnostic, tool-agnostic approach that allows contributors flexibility in tooling choice.
Reference implementations (such as Naveen's green code skill) will continue to be developed as a key mechanism for embedding patterns in real-world workflows.
Activity
[-][DRAFT] 2026-07-02 Green Software Principles & Patterns agenda[/-][+]2026-07-02 Green Software Principles & Patterns agenda[/+]pdematos12 commentedon Jul 2, 2026
I'm here
russelltrow commentedon Jul 2, 2026
Attended
navveenb commentedon Jul 2, 2026
Attended
babu-velu commentedon Jul 2, 2026
Attended
russelltrow commentedon Jul 2, 2026
Meeting Minutes
Summary
The Patterns Project held its regular working session on 2 July 2026, bringing together the project lead, operations leadership, and members from contributing organisations including AVEVA and Goldman Sachs. The team reviewed execution of the Patterns 2.0 initiative, validated the effectiveness of pattern proposal review processes, and advanced two critical pull requests through the approval pipeline. Naveen's nine AI patterns are progressing toward merge following incorporation of Green AI Committee feedback. Pratik's new Edge AI pattern has been routed for technical review by domain experts. The team also discussed strategic next steps for catalogue harmonisation, welcomed a new AVEVA contributor, and confirmed continued effectiveness of the pattern skill for proposal refinement.
Key Discussions
Pull Request #414 — Edge AI Pattern
Pratik from AVEVA submitted a pattern on optimizing power performance for Edge AI inference and training. The pattern was run through the pattern skill, which generated targeted feedback on scope, focus, and coverage areas. Pratik incorporated all feedback, and the pattern is now ready for formal approval. However, Liya raised an important quality gate: AI-domain patterns should be reviewed by the Green AI Committee to validate technical accuracy and domain expertise, not just structural compliance. Russ and Liya agreed to engage Goldman Sachs colleagues on the committee to arrange this review. Once technical sign-off is received, Liya will approve and merge. This review can be raised at the Green AI Committee's scheduled call next week.
Pull Request #407 — AI Patterns
Naveen has developed nine AI patterns as part of the foundational work on AI sustainability. The patterns received detailed feedback from Christian at Nuremberg University (Green AI Committee member), which Naveen has incorporated. Russ and Naveen discussed whether to seek further feedback before merge or proceed with an iterative approach. They agreed that Christian's feedback was comprehensive and substantive enough to proceed to merge, with the understanding that further iterations can happen post-merge if needed. Russ will review and close out all comments in the PR; once complete, Liya is ready to merge. Paula raised a design consideration: the nine patterns span multiple lifecycle phases (Architecture through Development), and merging a large batch at once could create category imbalance. The team acknowledged the risk and agreed to monitor for any need to create subcategories (e.g., splitting AI patterns by inference vs. training).
Pattern Submissions & Skill Effectiveness
Beyond the two PRs, the team reviewed incoming pattern proposals from multiple sources, including a new collection of green arrow patterns and additional AI-related patterns from AVEVA contributors. The pattern skill continues to demonstrate value: it catches common issues such as overly broad scope spanning multiple domains, missing sections (particularly cost impact), and insufficient grounding in user requirements. The team confirmed this validation process is working well and will remain standard for all future submissions.
Reference Implementations & Embedding Patterns in Tooling
https://github.com/Green-Software-Foundation/reference-implementations/tree/main/tools/gsf/green-code-skill
Naveen Balani shared a green code skill he has developed as part of the foundation's reference implementation work. This represents a key strategic direction: bringing patterns to life by embedding them into tools and workflows that developers actually use, rather than leaving them as standalone documentation. The team recognised this as important for adoption and discussed how to scale this approach across other pattern domains.
Catalogue Recategorisation & New Lifecycle Framework
The project has completed a major recategorisation of the patterns catalogue against a SWEBOK-based lifecycle framework (Requirements, Architecture, Design, Development, Testing, Operations, Decommissioning, End User). This restructuring has clarified where patterns fit within a complete software lifecycle and created more intentional homes for contributions. As a result, several patterns proposed in 2023 — which previously had no obvious category — now have clear placement. For example, architecture patterns were requested in 2023; the new Architecture category now provides that space. The team will revisit these older proposals to validate whether submitters still wish to contribute under the revised structure.
Catalogue Cleanup & Consolidation
Russ Trow has completed an assessment of proposed catalogue changes (deletions, renames, and consolidations) with detailed evaluation of validity and implementation approach. These recommendations will be reviewed offline by the project lead for approval, then prioritised at the next meeting. Alongside this, the team will triage the backlog of previously proposed but unresolved patterns to determine which should be advanced, which should be consolidated, and which are no longer relevant.
Design-Phase Tooling & Accessibility
The team discussed whether design-centric tools — specifically Claude Design and related AI-assisted design platforms — could support the creation and refinement of design-phase patterns. This opens the possibility of making pattern contribution more accessible to non-engineering roles (designers, product managers, UX practitioners) who may not be comfortable with current proposal workflows. The team noted that the design tools landscape is evolving rapidly and took an intentionally agnostic approach: rather than lock patterns into a specific tool, the goal is to develop a portable methodology or wrapper that contributors can apply using whatever tooling they prefer. This avoids vendor lock-in and maximises accessibility.
Contributor Engagement & Growth
A new contributor from AVEVA joined the project for the first time, bringing architectural expertise. The team welcomed broader engagement from member organisations, with several cross-organisational connections already forming (e.g., AVEVA contributors active on both the Patterns project and the SCI for Web initiative; GSF team members meeting AVEVA sustainability leads at external events like London Climate Action Week). The active submission of patterns from AVEVA (Pratik's Edge AI pattern, Babu's new participation, and broader engagement) suggests successful member recruitment.
Key Decisions
Next Meeting
Action Items