# Doc Store PG

> Mongo-compatible document workloads with search
> on PostgreSQL with CloudNativePG

A practical research program to test wether document storage can become 
a composable PostgreSQL capability -- and where MongoDB still wins.

Goal: turn architecture debate into reproducible evidence.

## Background

Initial Project for Research Tech Lead at Helix Forge.

## Reference Stack: one PostgresSQL substrate, multiple document surfaces

Measured compatibility, not assumed compatibility

### Mongo API Surface

Drivers connect through DocumentDB GW or FerretDB

### PostgreSQL document core

BSON / JSONB, document APIs, indexes, SQL hybrid queries

### CloudNativePG operations

HA lifecycle, extensions, databases, backup and restore

### The question

Can we preserve MongoDB developer ergonomics while gaining PostgreSQL's
relational model, extension ecosystem, and CloudNativePG operating model?

Measure the ladder:
MongoDB -> PostgreSQL JSONB -> DocumentDB APIs -> gateways

## Tech Lead Mission

Turn the hypothesis into a working lab, a benchmark discipline, and  
decision-grade evidence.

Build: reference deployments
Benchmark: workload ladder
Stress-test: tenant isolation ladder
Decide PostgreSQL vs. MongoDB trade-offs

### First milestone

Repeatable lab, first compatibility matrix, tenant-isolation result, written recommendation.


### Why this is attractive for a lead

It is small enough to build, deep enough to matter, and concrete enough to prove.

The outcome is not a benchmark blog post -- it is an operating model for document storage inside a PostgreSQL-first platform.

## Follow Up Potential

- Extend scope to include search capablility with PostgreSQL as an alternative to ElasticSearch.
- Utilize DocStorePG as the backend technology for markitect and derivative projects including coulomb.social.

xxx