# YelpArchive/MOE

A global, black box optimization engine for real world metric optimization.

Repository: https://github.com/YelpArchive/MOE
Canonical: https://ross.abutalabs.com/products/moe
Language: C++
License: NOASSERTION
License Family: other
Archived: true
Last push: 2023-03-24T11:00:32+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 4573, "days_push": 1258, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1321, forks 139 (observed 2026-08-28T04:04:21.889239+00:00)

## What it is
MOE (Metric Optimization Engine) is a global, black box optimization engine built by Yelp for optimizing expensive, time-consuming objective functions using Bayesian global optimization and optimal learning. It is written in C++ with a Python interface and REST API, and is ideal when function evaluations are costly, derivatives are unavailable, and a global optimum is sought.

## Use cases
- optimize click-through rate via A/B tests without wasting traffic
- tune hyperparameters of a slow machine-learning model
- optimize engineering designs that require expensive simulations
- optimize parameters of costly real-world lab experiments
- find global optima of non-convex black-box functions with no derivatives
- minimize the number of expensive objective function evaluations

## When to choose
- each evaluation of your objective function is expensive or slow (A/B tests, simulations, lab experiments)
- the objective is a black box with no gradients and possibly non-convex
- you want Bayesian optimization with historical data support
- you need a REST API for an optimization service

## When to avoid
- your objective function is cheap to evaluate (use standard gradient-based or scipy optimizers)
- you need gradients or convex optimization guarantees
- you need actively maintained software with up-to-date dependencies (CI is broken, last release 2023)
- you need a pure-Python lightweight library

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, benchmarking
- domain: machine-learning, data-science
- platform: python, cpp
- tags: bayesian-optimization, black-box-optimization, optimal-learning, hyperparameter-optimization, rest-api, algorithms, docker, linux, macos

## Member repositories
- YelpArchive/MOE (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:21.889239+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:47:16.523745+00:00, confidence not recorded.
  - readme: https://github.com/YelpArchive/MOE (fetched 2026-08-28T04:04:21.889239+00:00, sha c0a9d50bca15)
- Data as of 2026-08-30T08:39:29.467469+00:00.