<!-- Generated from the rendered page by scripts/write-llm-mirrors.mjs. Do not edit by hand. -->
Canonical: https://molo17.com/blog/coming-in-gluesync-2-3-ai-powered-data-integration/
Markdown mirror: https://molo17.com/blog/coming-in-gluesync-2-3-ai-powered-data-integration/index.md
Title: AI Hub is coming — not another chatbot on your sync tool | MOLO17
Description: AI Hub is coming with Gluesync 2.3 — not another chatbot on your sync tool. Book a demo or request early access to the beta.

[← All articles](/blog/)

Article

# AI Hub is coming — not another chatbot on your sync tool

![AI Hub is coming — not another chatbot on your sync tool featured image](/blog/coming-in-gluesync-2-3-ai-powered-data-integration/27bfc053-image.jpg)

13 Sep 2026  [Gluesync](/blog/?category=gluesync)  2 min read

AI Hub is coming this month — not another chatbot on your sync tool. Gluesync 2.3 puts AI in the control plane across 59+ databases.

This is how we envisioned AI-powered data integration.

Bringing AI to 59+ databases — including legacy systems, no matter how old.

That is the line for **[Gluesync 2.3](https://molo17.com/ai-hub/)**. Not a side chatbot bolted onto a sync tool. AI that sits in the control plane you already operate, under the permissions you already grant, against the agents and schemas you already trust.

## Why 2.3 is about AI in the Hub

Most “AI for data” demos assume a clean warehouse and a shared service account. Real pipelines do not.

You have Oracle next to Postgres next to something that still speaks journal receivers. You have RBAC that actually matters. You have operators who will not accept a model that invents SQL and runs it because a prompt looked confident.

Gluesync 2.3 puts AI where those constraints already live: **Core Hub**.

## What is coming

### [AI Studio](https://molo17.com/capabilities/ai-studio/)

A place in Core Hub to work with agents that know your pipelines and can ask your data — with the model **you** bring.

BYO LLM in Hub settings (your keys, your providers). Tool access stays inside Hub RBAC and allow-lists. Traces stay visible. You shape the agent; you do not outsource the permission model.

### [Query Studio AI](https://molo17.com/query-studio/)

Natural language in. Proposed SQL out. Same editor, same schema, same permissions.

The helper drafts. **You** review before anything runs. It is not an auto-execute path. Grounding is the open agent and the catalog you already use in Query Studio.

### [Core Hub MCP](https://molo17.com/corehub-mcp/)

Expose permission-aware Hub tools to approved AI clients over MCP. Same brain the rest of Gluesync AI leans on: pipelines, entities, metrics, governed SQL helpers — under the caller’s identity, not a shared god account.

If you already use Core Hub MCP today, 2.3 is the broader AI surface built on that foundation.

### SQL-callable AI on [Query Forge](https://molo17.com/query-forge/#sql-callable-ai)

For teams that live in JDBC clients, 2.3 extends the idea further: ask from SQL against a virtual catalog such as [`gluesync.ai`](https://molo17.com/query-forge/#sql-callable-ai), and get an answer path that still respects Hub RBAC and does not silently run proposed statements for you.

Same Forge connection style you already know (`jdbc:gluesync://…`).

### [AI on legacy databases](https://molo17.com/solutions/ai-on-legacy/)

The point of “59+” is not a logo wall. It is that the AI path should reach the same agents Gluesync already syncs — including systems that predate cloud-native everything.

If the data still matters in production, the AI surface should be able to meet it there, under the same identity and safety gates.

## What does not change

-   Your infrastructure stays yours.
-   Your models stay yours (BYO).
-   Your permissions stay yours.
-   Drafts and proposals are not a free pass past SafetyGate / RBAC.

We are not asking you to trust a black-box “data AI” that bypasses the Hub. We are putting AI _inside_ the Hub.

## What’s next

**[AI Hub](https://molo17.com/ai-hub/) is coming this month** — sooner than you might think.

Want a sneak peek of what’s coming? [Book a demo](https://molo17.com/book-a-demo/) with us today, or request early access to the beta.

If you want the short version: **Gluesync 2.3 is AI-powered data integration across the databases you already run — including the old ones — without giving up the control plane.**

[← Older article Gluesync Core Hub MCP server: pipelines, SQL, and ops from your AI client](/blog/gluesync-core-hub-mcp-server/) [Newer article → Gluesync Couchbase DCP source agent: native CDC without Eventing](/blog/gluesync-couchbase-dcp-source-agent/)

## Keep reading

1.  [Gluesync Query Forge: cross-database federated SQL querying without data movement Query Forge turns Gluesync Core Hub into a secure federated SQL gateway. Run cross-database JOINs across remote agents using a dedicated JDBC driver with zero data movement. 5 Aug 2026](/blog/query-forge-federated-sql-querying/)
2.  [Gluesync Gluesync 2.2.10: federated SQL querying with Query Forge, advanced scheduling in Chronos and redesigned Control Plane Gluesync 2.2.10 transforms real-time replication into a unified data platform. Discover Query Forge federated SQL, Chronos event chaining, new UI and security enhancements. 28 Jul 2026](/blog/gluesync-2-2-10-query-forge-chronos-pipeline-wizard-ui/)
3.  [Gluesync Gluesync 2.2.9: Query Studio, Universal File Store Agent, Filed Functions and more 2.2.9 introduces Query Studio, new Control Plane UI, native DB2 LUW log-based CDC, Field Functions and the iOS/Android mobile companion app. 30 Jun 2026](/blog/gluesync-2-2-9/)

[Back to all articles](/blog/) [More in Gluesync →](/blog/?category=gluesync)
