v1.0.0 Roadmap

[jjk-g]

P0

• [Feature] Multi-Modal Inference Benchmarking
• [Feature] HF OTel trace loading (IBM OSS)
• [Testing] Identify Testing gaps and create test plan
• [UX] Image & Helm Chart published in k8s Add inference-perf promoter manifests kubernetes/k8s.io#9267
• [UX] User Guides & Docs
• [Metrics] Standard Report Format
• [Release] 1.0.0 Blog Post
• [Release] Release Cadence Definition

P1

• [UX] Provide several agent skills that apply our best practices and common patterns
• [Process] Added Repo roles for active contributors
• [UX] Feature parity in llm-d-benchmark

P2

• [Loadgen] Distributed Loadgen (Inference-perf-engine; Production Scale Loadgen)
• [Metrics] GPU utilization / other hardware metrics from Prometheus

[jjk-g]

Notes from 1/22/26 meeting

• Added functionality to analysis module (multiple runs on same graph, benchmark results summary/analysis)
• Review test plan with @huaxig and code coverage. + code coverage presubmit.
• Review crucial bug fixes and add (concurrency issue)
• Image + Helm chart publishing

[Bslabe123]

@jjk-g given this has gone stale, lets close this one, sounds good?