Frequently asked questions.

Insight O' Mate is a privacy-first natural language database querying platform. Type a question in plain English and it generates the corresponding MongoDB, Firestore, Redis, or DynamoDB query — then runs it locally on your machine and returns the results.

Four NoSQL databases today: MongoDB, Redis, Firestore, and DynamoDB. More NoSQL adapters (Cassandra, Elasticsearch) are on the roadmap; see integrations for the full, current list.

A database GUI (like MongoDB Compass) still requires you to write queries manually. Insight O' Mate generates the query from your plain-English question — no MQL, no Firestore query syntax, no Redis command to memorise.

Yes. Type a question like “Show me the top 10 customers by total orders this month” and Insight O' Mate translates it into a MongoDB aggregation pipeline, runs it, and returns the results. See the MongoDB page for examples.

Yes. Connect your Firebase project and ask questions in plain English. Insight O' Mate generates the Firestore collection query and returns the matching documents. See the Firestore page for examples.

Yes. Ask about your Redis keys, hashes, sets, sorted sets, and streams in plain English. See the Redis page for examples.

Yes. Insight O' Mate translates questions into DynamoDB GetItem, Query, and Scan operations — handling key conditions and filter expressions automatically. See the DynamoDB page for examples.

No. Your documents, field values, and connection string stay on your computer. The only thing sent to our NLP engine is your prompt plus your field and collection names — never the values inside them. See the security page for exactly what travels and how to verify it.

In your operating system's keychain — macOS Keychain, Windows Credential Vault, or libsecret on Linux — with an AES-256-GCM encrypted-file fallback. It's never written to disk in plaintext and never transmitted.

No. The hosted engine is stateless: it reads a prompt, returns a structured analysis, and forgets both. There's no prompt database, no input logging, and no training feedback loop on your data.

Currently, query generation requires our stateless cloud engine. However, we plan to offer a self-hosted VPC engine in the future for enterprises that need to run the NLP engine fully offline.

Yes — the free plan is genuinely free, with a daily query quota. Paid plans add higher limits, query history, templates, and team features. See pricing for the full breakdown.

Each prompt you send to the NLP engine counts once against your quota, measured after the call returns. Re-running a saved result against your live database does not re-bill the engine.

Yes. Plans are managed from your dashboard and you can cancel at any time; you keep access through the end of the current billing period.

Four NoSQL databases (MongoDB, Redis, Firestore, and DynamoDB) plus Excel and CSV spreadsheets today. More NoSQL adapters (Cassandra, Elasticsearch) are on the roadmap; see integrations for the full, current list.

macOS, Windows, and Linux (AppImage and .deb). Grab the build for your OS from the download page.

It's strongly recommended. A read-only role means an accidental write operation from a misread prompt can't mutate your data. The docs walk through creating one for Atlas and self-hosted MongoDB.

Create one from your dashboard once you've signed in. The desktop app uses it to authenticate to the NLP engine and to resolve your plan and quota.

Reach us from the contact page, or use in-app support once you're signed in. Check status first if something seems down.