# rentruewang/aioway

AI on the way. An auto deep learning pipe dream. An RDBMS approach to deep learning. Declarative, explainable, scalable, optimizable, easy to deploy, all that good stuff.

Repository: https://github.com/rentruewang/aioway
Canonical: https://ross.abutalabs.com/products/aioway
Homepage: https://aioway.rentruewang.com/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: pytorch, lazy-evaluation, python, compiler, neural-architecture-search, relational-algebra, explainable-ai, sql, auto-deep-learning
Last push: 2026-08-26T18:46:12+00:00

## Health v2 (maintenance only)
Score: 77/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 35, longevity 100
- inputs: {"age_days": 1750, "days_push": 7, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1824, forks 65 (observed 2026-08-28T04:05:41.228014+00:00)

## What it is
Aioway is an optimizing compiler for deep learning algorithms that treats ML models as instructions and builds declarative pipelines using relational algebra (SQL-like) and Python interfaces. It aims to auto-select models and algorithms based on tasks and resources while keeping models explainable and scalable.

## Use cases
- automatically build deep learning pipelines without expert tuning
- declaratively define ML workflows with SQL-like syntax
- select the best model for my task and hardware automatically
- get explainable white-box ML models instead of black-box AutoML
- scale up model size and migrate training across machines
- extend an AutoML pipeline with custom PyTorch models

## When to choose
- you want a fast, rule-based alternative to slow neural architecture search
- you need explainable, declarative ML pipelines with lazy evaluation
- you want to mix relational/SQL-style queries with PyTorch models
- you find existing AutoML tools too inflexible or black-box

## When to avoid
- you need a mature, production-hardened AutoML framework (pre-v0.1.0)
- you want a UI-driven no-code AutoML experience
- you need pretrained models or LLM inference out of the box
- you require a large community and extensive documentation

## Facets
- artifact type: library
- maturity: experimental
- function: compiler, machine-learning, deep-learning, llm-training, parser
- domain: deep-learning, machine-learning, artificial-intelligence, developer-tools
- platform: python, cross-platform
- tags: automl, relational-algebra, sql, lazy-evaluation, neural-architecture-search, explainable-ai, pytorch, declarative-pipeline

## Member repositories
- rentruewang/aioway (main) score 77

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:41.228014+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 2026-08-30T03:19:44.257737+00:00, confidence not recorded.
  - readme: https://github.com/rentruewang/aioway (fetched 2026-08-28T04:05:41.228014+00:00, sha 652eeeb523d7)
  - homepage: https://aioway.rentruewang.com/ (fetched 2026-08-29T10:58:58.256643+00:00, sha 44136fa355b3)
- Data as of 2026-08-30T08:39:29.467469+00:00.