Agentlytics
MongoDB

Query MongoDB in plain English. Aggregations and dashboards without writing pipelines.

Connect MongoDB or MongoDB Atlas with a read-only user and Agentlytics samples your collections to learn their fields. Ask a question in plain English and it builds a validated aggregation pipeline, runs it live, and turns the result into charts and dashboards you can share.

shop (MongoDB Atlas)Example

Orders per day by status for the last 14 days

Delivered orders averaged 1,240 per day; cancellations spiked to 6.1% last Tuesday.
orders collection - aggregation pipeline
Example question and answer for the MongoDB AI Analyzer

What the MongoDB AI Analyzer does

Purpose-built AI agents and deterministic tools, so answers about your data are computed and cited - not guessed.

Safe aggregation pipelines

Write stages such as $out and $merge, and server-side JavaScript operators, are blocked, and every query has a row limit and timeout.

Learns your collections

Fields and types are inferred from sampled documents, including nested fields, so pipelines use the right paths.

Works with Atlas

Connect a self-hosted MongoDB or an Atlas cluster with a standard or SRV connection string.

Live dashboards

Dashboards store the pipeline, not your documents, and re-run it on open with relative time windows.

Encrypted credentials

Connection strings are encrypted, never shown in the browser and never logged.

Retries with context

If a pipeline fails, the error is fed back so the AI can correct the query before showing you a result.

How it works

  1. 01

    Create a read-only user

    Use the in-app guide to create a user with the read role on your database.

  2. 02

    Connect your cluster

    Paste the connection string, test it and review the detected collections.

  3. 03

    Ask for answers or dashboards

    Ask in plain English and get charts backed by live aggregation pipelines.

Questions you can ask

Type them exactly like this - no formulas, SQL or setup.

  • How many new users signed up each week?
  • Top 5 product categories by order count
  • Average session length by device type
  • Build a dashboard of orders, revenue and cancellations
  • Which customers have more than 3 open tickets?
  • Revenue by country for the last 30 days

Popular use cases

Teams use the MongoDB AI Analyzer for work like this.

App analytics

Analyze users, events and sessions stored as documents.

E-commerce

Track orders, carts, refunds and product performance live.

Support & ops

Monitor tickets, statuses and response times across collections.

IoT & events

Aggregate high-volume event documents into daily trends.

MongoDB AI Analyzer FAQ

Does it work with MongoDB Atlas?

Yes. Connect an Atlas cluster or a self-hosted MongoDB using a standard mongodb:// or mongodb+srv:// connection string.

Can the AI change or delete my MongoDB data?

It is designed not to. Connect with a read-only user, and pipelines that write data or run server-side JavaScript are rejected before they reach your database.

Are my documents stored by Agentlytics?

No. Only the pipeline and chart settings are saved. Results are fetched live when you open a dashboard.

Do I need to write aggregation pipelines?

No. Describe what you want to see and the pipeline is generated from your collection's fields, validated, and run for you.

Try the MongoDB AI Analyzer free

Create a workspace, add your data and ask your first question in under a minute.