Skip to content

Enterprise adoption evaluation

This page summarizes what a documentation-only evaluator can conclude about OntoSQL 0.5.x for large-organization adoption. It is honest about beta status and known documentation limits.

0.5.x beta

OntoSQL is not GA. Semver guarantees begin at 1.0. Pin versions and complete the evaluation checklist before production use.

Overall verdict

OntoSQL is documented honestly as pre-1.0 beta software with strong architecture and security transparency. It is suitable for controlled PoC (Postgres, pinned version, SQL-first team). Enterprise platform standardization should wait for 1.0, a support policy, and performance evidence.

What does OntoSQL do?

OntoSQL is a Python semantic data access layer for SQL-first applications:

Layer Tool Role
Physical SQLModel (table=True) Mirrors real DB tables
Semantic OntoModel Application entities, validation, ontology metadata
Mapping OntoMapper / Map Explicit field → column/join bindings
Runtime OntoSession / AsyncOntoSession Compiles get, find, save, delete to SQL

Optional extras: JSON-LD/RDF export, RDF import, post-commit graph sync, SHACL, FastAPI OntoRouter.

Not: SPARQL database, OBDA engine, OWL reasoner, or automatic table→ontology inference. See When to use OntoSQL.

Why would I use it?

Need OntoSQL value
Legacy/normalized SQL + ontology-shaped Python APIs Explicit maps over joins, bridges, computed fields
Optional JSON-LD/RDF/FastAPI negotiation Export from mapper metadata
Hybrid SQL system-of-record + RDF mirror Post-commit graph_sync — HYBRID.md
Multiple semantic views of one table Multi-map views

Cost: maintain row models, semantic models, and mappers (three artifacts).

Is it production ready?

No — and the docs say so.

Signal Source
PyPI Development Status :: 4 - Beta COMPATIBILITY.md
Semver from 1.0 only; pre-1.0 may rename without deprecation SPECS.md
OntoRouter is a scaffold — auth/authz required SECURITY.md
Graph sync is eventually consistent after SQL commit HYBRID.md
Bulk write, read replicas, OTel Planned 0.7
GA / Production/Stable classifier Planned 1.0

Enterprise interpretation: PoC with pinned versions — yes. Platform standard — wait for 1.0 + support policy.

Is it actively maintained?

Partially documented.

Known Unknown from docs
Release cadence in changelog (0.2.0 → 0.5.x) Maintainer team size, bus factor
CI badge, contributing workflow Issue/commit response SLAs
Public roadmap through 1.0 Calendar date for 1.0
Read the Docs hosting Commercial backing or LTS windows (until 1.0)

See SUPPORT.md for current community support model.

How difficult is migration?

Scenario Difficulty Documentation
Greenfield app Moderate Quick start
Existing SQLAlchemy/SQLModel app Moderate–high Manual maps; Alembic guide
0.4.x → 0.5.x Low (additive) Changelog migration section
0.2.x → current / pre-1.0 → 1.0 High Upgrading guide (partial; full notes at 0.9)

Migration is mapper authoring + Alembic coordination, not a drop-in ORM swap.

How difficult is onboarding?

Persona Difficulty Doc support
Python dev, SQLModel familiarity Low–moderate Quick start, semantic queries
Async / FastAPI Moderate async, FastAPI quick start
Hybrid SQL + RDF High HYBRID.md, graph sync operations
Enterprise platform / security High This page + compliance guide

Time-to-first-success (typical): Tier 1 CRUD 5–15 minutes; production FastAPI hours–days; hybrid graph sync days–weeks.

What are the risks?

Documented technical risks

Risk Severity Read
Pre-1.0 API instability High SPECS, COMPATIBILITY
CascadePolicy.REPLACE data loss High cascade policies, SECURITY
Graph/SQL split-brain High HYBRID, graph sync ops
OntoRouter without auth High production-router
RDF import DoS / PyLD SSRF Medium SECURITY
TripleModel core dependency (SQL-only apps) Medium COMPATIBILITY

Enterprise / procurement risks

No vendor SLA, performance benchmarks, compliance certifications, or production case studies today. See compliance guide and SUPPORT.md.

Adoption recommendation

Scenario Recommendation
Enterprise platform standard Do not adopt until 1.0 + support policy + benchmarks
Team PoC (Postgres, pinned 0.5.x) Proceed with spike — budget mapper work and security review
Hybrid SQL + RDF production High caution — custom reconciliation; see graph sync ops
Public OntoRouter Reject default scaffold — custom auth/authz only

Evaluation checklist

Use this before approving OntoSQL beyond a time-boxed PoC.

Fit and scope

  • [ ] Read When to use OntoSQL — confirm you need semantic layer, not raw SQLModel or SparqlModel
  • [ ] Confirm Postgres or SQLite (only CI-tested DBs) — COMPATIBILITY
  • [ ] Accept 0.5.x beta and pin ontosql==… in requirements
  • [ ] Budget mapper authoring for every entity (no auto-inference)

Security and API

  • [ ] If exposing HTTP: do not mount OntoRouter without dependencies=[Depends(auth)] — SECURITY
  • [ ] Object-level authorization designed for {entity_id} and nested POST/PATCH bodies
  • [ ] Rate limits at proxy or middleware
  • [ ] RDF import endpoints authenticated; untrusted=True and byte/triple caps
  • [ ] Review compliance guide for questionnaire gaps

Data and operations

  • [ ] Alembic owns physical schema; mappers updated with column changes — Alembic guide
  • [ ] Cascade policy per nested field reviewed (link default; avoid REPLACE on shared rows)
  • [ ] If graph_sync: split-brain runbook, monitoring, retry_graph_sync — graph sync ops
  • [ ] SQLite production: PRAGMA foreign_keys=ON if using REPLACE cascade

Quality and upgrades

Still open (documented gaps)

These cannot be signed off from documentation alone today:

Gap Status
Support / LTS policy Pre-1.0 — SUPPORT.md
Performance benchmarks Planned 0.9 — ROADMAP
Production case study Planned 1.0
MySQL / Oracle / SQL Server certification Not CI-tested
SOC2 / HIPAA / FedRAMP mapping Not certified — compliance

Summary

Question Answer from docs
What does it do? SQL-first semantic CRUD + optional RDF/FastAPI
Why use it? Legacy SQL + ontology APIs, hybrid graph, content negotiation
Production ready? No — 0.5.x beta
Actively maintained? Yes (OSS); no commercial SLA
Migration difficulty? Moderate–high; 0.4→0.5 documented
Onboarding difficulty? Moderate
Risks? Technical risks documented; enterprise/compliance gaps remain

Read next: SUPPORT.md · Compliance · Security · Compatibility