Skip to content

CI: reproducible benchmark harness with tracked results #435

Description

@HarshaNalluru

Description

A benchmark suite that runs in CI and records latency and recall over time.

Use case

The project wants to catch performance regressions and publish numbers people can trust.

Proposed solution

Add a harness that runs a fixed dataset in CI and stores p50/p95/p99 and recall per commit, with a guard against regressions. Benchmark reference: https://docs.moss.dev/docs/start/what-is-moss Contributing guide: https://github.com/usemoss/moss/blob/main/CONTRIBUTING.md

Alternatives considered

No response

Metadata

Metadata

Assignees

No one assigned

    Labels

    help wantedExtra attention is neededtoolingCLI, dev tooling, and CI

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions