Operations and business teams
Today: Answers live in three systems and nobody can join them up.
With the Data Lake: One governed database holds ERP, CRM and e-commerce data together, so reports and internal tools finally see the whole picture.
Overview
What this is,
without the jargon
This page is for anyone deciding whether Commercient Data Lake \ Data Hub is worth a look. It explains what the product does, what the command-line tool is for, and defines every term you will meet on the rest of the site. No commands required.
One platform
Bring your ERP, CRM, e-commerce and other data together in a single governed database.
Ready-to-use APIs
Choose which tables to publish and REST and GraphQL endpoints exist for them immediately.
Built for AI agents
An agent connects with a revocable key and only ever sees the tables and fields you allow.
The problem it solves
Most companies run their business across several systems: an ERP for finance and operations, a CRM for sales, perhaps an e-commerce platform, and a few databases besides. Each holds part of the truth. None holds all of it.
So every new report, internal tool or AI project starts the same way. Someone has to get to the data: request access, write a pipeline to copy it somewhere, build an interface for applications to read it, then work out who is allowed to see what. That groundwork takes weeks, and it is redone for the next project.
What the product does instead
Commercient Data Lake \ Data Hub gives your company its own managed database, built on Microsoft SQL Server, kept current from the systems you already run. On top of it sit the three ready-made ways in above: one platform for your data, APIs for your applications, and governed access for AI agents.
Because it is a real SQL Server database underneath, the guarantees are database guarantees. Row rules, auditing, history and permissions are native features rather than promises made by an application layer, and your team's existing SQL skills carry straight over.
And the CLI?
In plain wordsA command line is a way to tell software what to do by typing instructions instead of clicking. dlake is that for the Data Lake.
Everything the web console does, dlake can do from a terminal: create the Data Lake itself, bring data in, define tables, publish APIs, hand out and revoke keys, run queries and export files. Because it is typed rather than clicked, the same sequence can be saved, reviewed and run again next month or on another customer's tenant.
That matters for three reasons. Routine work stops depending on one person remembering the order of steps. Set-up becomes repeatable across customers. And an AI assistant can drive the same commands, with the same permissions, under the same audit trail.
It is one self-contained program for Windows, macOS and Linux, and it is free to use with a Commercient subscription.
# create the Data Lake
$ dlake register start --email you@acme.com
# define a table and publish it as an API
$ dlake admin create_table \
--tableName orders --columns @orders.json
$ dlake admin set_entity_exposure \
--entity orders --expose true
$ dlake admin restart_dab --confirm true
# a key that can only read that table
$ dlake admin create_api_key --keyName web
Who gets value from it
IT and developers
Today: Every project starts by building a pipeline and an API layer.
With the Data Lake: The database, the APIs, the keys and the audit trail already exist. You define tables and publish them.
Teams adopting AI
Today: An assistant is only as useful as the data it can reach, and giving it access feels risky.
With the Data Lake: An agent connects with a revocable key limited to named tables and fields, enforced by the database and recorded in the audit log.
Existing Commercient SYNC customers
Today: Data already flows between the ERP and the CRM, but nothing else can use it.
With the Data Lake: The Data Lake adds an API and an AI layer over the same connected data, without changing the integration you already run.
The words you will meet
In plain wordsTen terms that appear across this site and inside the product, each in one sentence.
- Tenant
- Your company’s own Data Lake: one managed SQL Server database, separate from every other customer’s.
- Schema
- The shape of your data: which tables exist, what columns they have, and how they relate.
- API
- A way for one piece of software to ask another for data. Your apps use it instead of talking to the database directly.
- REST and GraphQL
- The two standard styles of web API. The platform creates both for the tables you choose to publish.
- API key
- The secret an app or an AI agent presents to prove who it is. A key can be limited to certain tables and actions, and revoked at any time.
- MCP
- Model Context Protocol: the open standard that lets an AI assistant use tools and data safely.
- Row-level security
- Rules inside the database deciding which rows each caller may see, applied on every read whoever is asking.
- CLI
- Command-line interface: a program you drive by typing commands, so the work can be scripted and repeated.
- Data API engine
- The service that turns your published tables into web APIs. Some changes reach it only after you restart it.
- Commercient SYNC
- Commercient’s integration product, which moves data between more than 150 business systems.
