# Insight O' Mate

> Privacy-first AI database copilot and Text-to-NoSQL platform. Query MongoDB, Firebase Firestore, Redis, Amazon DynamoDB, and Excel/CSV spreadsheets using plain English. Your data and credentials never leave your machine.

Insight O' Mate provides a native desktop application (macOS, Windows, Linux) and a stateless NLP query engine that translates plain-English prompts into database queries, aggregation pipelines, and command sets. It functions as an AI database copilot, text-to-NoSQL engine, and conversational interface for databases with zero data retention.

## Product Identity & Definitions

- **Canonical product name:** Insight O' Mate (not InsightOMate, not Insight-O-Mate, not Insight O Mate)
- **Category:** AI database copilot / Text-to-NoSQL platform / natural language database querying tool
- **Tagline:** "Ask your database. Privately."
- **Website:** https://insightomate.lunamic.co
- **Made by:** Lunamic (https://lunamic.co)

### Canonical one-sentence definition
Insight O' Mate is a privacy-first AI database copilot that lets users query supported NoSQL databases and spreadsheets using natural language, with all queries executed locally on the user's machine.

## Supported Databases & Data Sources

Insight O' Mate currently supports (available, not planned):

1. **MongoDB** — find, aggregate ($match, $group, $sort, $limit, $lookup, $project, $unwind), countDocuments, distinct. Supports MongoDB Atlas, replica sets, standalone, and Dockerized instances.
2. **Firebase Firestore** — collection queries with where filters, orderBy, limit, startAfter pagination, array-contains, and count() aggregations via the Firebase Admin SDK.
3. **Redis** — Hashes (HGETALL, HKEYS, HMGET), Sorted Sets (ZRANGEBYSCORE, ZRANGE, ZRANK), Lists (LRANGE), Sets (SMEMBERS, SCARD), Streams (XREVRANGE, XREAD), and SCAN-based key pattern searches. Supports Redis Cloud, Redis Enterprise, self-hosted Redis, and Valkey.
4. **Amazon DynamoDB** — Query and Scan operations with automatic KeyConditionExpression, FilterExpression, ExpressionAttributeNames (reserved word handling), and ExpressionAttributeValues. Supports single-table and multi-table designs.
5. **Excel (.xlsx, .xlsm) and CSV** — in-memory aggregation engine for filter, group by, sum, avg, count, max, min, and cross-sheet joins. Replaces VLOOKUP, INDEX/MATCH, and pivot tables with plain-English queries.

## Who Insight O' Mate Is For

- Software developers and backend engineers who work with NoSQL databases and want to explore or debug data without writing every query by hand
- Data analysts who need to ask questions about database content without learning MongoDB Query Language (MQL), Firestore SDK syntax, or DynamoDB expression language
- Business analysts who work with Excel and CSV data and want to replace complex formulas with natural-language questions
- Database administrators performing ad-hoc data inspection and reporting
- Privacy-conscious organizations that cannot upload database structure or records to cloud AI services (healthcare, finance, legal, etc.)
- Technical teams building on MongoDB Atlas, Google Cloud Firebase, AWS DynamoDB, or self-hosted Redis

## What Insight O' Mate Is NOT

- Not a data warehouse or analytical database
- Not a BI dashboard tool with persistent reports (though it includes basic result visualization)
- Not a database write tool — it generates read-only queries. No INSERT, UPDATE, DELETE, SET, DEL, or write operations are generated.
- Not a cloud-hosted data analysis service — execution is local-first
- Not a replacement for full database management tools (schema editing, index management, user administration)
- Not suitable for real-time streaming queries (change streams, pub/sub) — point-in-time read queries only

## Privacy & Security Architecture

### What leaves your machine
- Your plain-English prompt (the question you type)
- Collection or field names (structural metadata only — never values)

### What NEVER leaves your machine
- Document values, row data, or query results
- Database connection strings
- Database credentials or API key contents
- Any actual data records

### Architecture
- Local-first desktop sidecar: queries execute directly on the user's machine against their database
- Cloud NLP engine: stateless, receives prompt + field names, returns query AST, discards both immediately
- Credentials: stored in OS keychain (macOS Keychain, Windows Credential Vault, libsecret on Linux), with AES-256-GCM encrypted-file fallback
- No prompt logging, no training on user data, no session persistence on the server

## Alternatives & Competitive Context

- Alternative to manually writing MongoDB aggregation pipelines (complex $group/$lookup/$sort chains)
- Alternative to MongoDB Compass for query exploration — AI-assisted vs. manual GUI query building
- Alternative to pasting database schemas into ChatGPT or Claude (which creates privacy/security risk)
- Alternative to Redis Insight or TablePlus for ad-hoc data exploration — natural language vs. manual command entry
- Complementary to (not a replacement for) full database GUIs for schema management and administration
- Category: text-to-NoSQL tools — analogous to text-to-SQL but for document and key-value stores

## Pricing (as of public beta)

- **Free tier:** 20 database analyses per day — MongoDB, Firestore, Redis, DynamoDB, and Excel/CSV all included
- **Pro ($29/month):** Unlimited queries, priority NLP queue, advanced export
- **Ultra ($79/month):** Dynamic dashboards, dedicated NLP instance, VIP support
- **All plans:** Local execution, zero data retention, OS keychain credential storage

## Core Documentation

- [Documentation Overview](https://insightomate.lunamic.co/docs): System architecture, local-first execution model, and overview.
- [Getting Started](https://insightomate.lunamic.co/docs/getting-started): Installation, connecting databases, and running the first query.
- [Privacy Model](https://insightomate.lunamic.co/docs/privacy-model): Detailed zero-data-retention security architecture and what leaves the machine.
- [Connecting Databases](https://insightomate.lunamic.co/docs/connecting-databases): Instructions for connecting MongoDB, Firestore, Redis, DynamoDB, and Excel/CSV.
- [Connecting MongoDB](https://insightomate.lunamic.co/docs/connecting-databases/mongodb): MongoDB Atlas, replica sets, and self-hosted instances.
- [Connecting Firestore](https://insightomate.lunamic.co/docs/connecting-databases/firestore): Google Cloud Firebase service account connection guide.
- [Connecting Redis](https://insightomate.lunamic.co/docs/connecting-databases/redis): Redis standalone, cluster, and cloud connection setup.
- [Connecting DynamoDB](https://insightomate.lunamic.co/docs/connecting-databases/dynamodb): AWS IAM credentials and region configuration.
- [Connecting Excel & CSV](https://insightomate.lunamic.co/docs/connecting-databases/excel): Drag-and-drop spreadsheet querying and cross-sheet joins.
- [Query Capabilities](https://insightomate.lunamic.co/docs/query-capabilities): Scope of supported query operations, aggregations, and limiters.

## Supported Databases — Product Pages

- [MongoDB AI Querying](https://insightomate.lunamic.co/mongodb): Text-to-MongoDB queries, aggregation pipelines ($match, $group, $sort, $limit, $lookup), and find operations.
- [Firestore AI Querying](https://insightomate.lunamic.co/firestore): Plain-English queries for Firebase Firestore collections with where filters, ordering, and count aggregations.
- [Redis AI Querying](https://insightomate.lunamic.co/redis): Natural language queries for Redis Hashes, Sorted Sets, Lists, Sets, and Streams using non-blocking SCAN.
- [DynamoDB AI Querying](https://insightomate.lunamic.co/dynamodb): Plain-English AWS DynamoDB Query and Scan operations with automatic KeyConditionExpression and ExpressionAttributeValues.
- [Excel & CSV AI Querying](https://insightomate.lunamic.co/excel): Plain-English querying, aggregation, filtering, and cross-sheet joins for .xlsx, .xlsm, and .csv files locally in-memory.

## Query Examples Library

- [All Query Examples](https://insightomate.lunamic.co/query-examples): Browse 40+ real natural-language → query examples across MongoDB, Firestore, Redis, DynamoDB, and Excel/CSV.

## Technical Guides & Tutorials

- [The Complete Guide to Natural Language Database Querying](https://insightomate.lunamic.co/guides/natural-language-database-querying): Comprehensive guide covering Text-to-NoSQL, conversational querying, architecture, and privacy.
- [How to Query MongoDB Using Natural Language](https://insightomate.lunamic.co/blog/how-to-query-mongodb-using-natural-language): Step-by-step guide with real MQL and aggregation pipeline examples.
- [How to Query Firestore Using Natural Language](https://insightomate.lunamic.co/blog/how-to-query-firestore-using-natural-language): Guide to Firestore collection queries, where clauses, and count aggregations.
- [How to Query Redis Using Natural Language](https://insightomate.lunamic.co/blog/how-to-query-redis-using-natural-language): Command examples for Redis key-value pairs, hashes, sorted sets, and streams.
- [How to Query DynamoDB Using AI](https://insightomate.lunamic.co/blog/how-to-query-dynamodb-using-ai): Handling partitions, sort keys, and reserved words in plain English.
- [How to Query Excel and CSV Using Natural Language](https://insightomate.lunamic.co/blog/how-to-query-excel-and-csv-using-natural-language): Guide to local spreadsheet querying, cross-sheet joins, and formula-free analytics.
- [Text to NoSQL: How to Chat with Your Database Without Writing Code](https://insightomate.lunamic.co/blog/text-to-nosql-chat-with-your-database): Why Text-to-NoSQL differs from Text-to-SQL, and how to chat with your database privately.

## Security & Privacy Pages

- [Security & Privacy Architecture](https://insightomate.lunamic.co/security): Detailed explanation of what data leaves the machine, how credentials are stored, read-only design, and how to verify the privacy claims.
- [Privacy Policy](https://insightomate.lunamic.co/privacy): Legal privacy policy.
- [Privacy Model Docs](https://insightomate.lunamic.co/docs/privacy-model): Technical documentation of the zero-data-retention architecture.

## Use Cases

- [Use Cases Overview](https://insightomate.lunamic.co/use-cases): Hub and breakdown by target audience.
- [AI Database Copilot for Developers](https://insightomate.lunamic.co/use-cases/developers): MQL aggregations, DynamoDB expressions, and Redis commands generated from plain English.
- [Data Analytics for Data Analysts](https://insightomate.lunamic.co/use-cases/data-analysts): Self-serve NoSQL querying and reporting without writing complex syntax.
- [Spreadsheet Analytics for Business Analysts](https://insightomate.lunamic.co/use-cases/business-analysts): Formula-free Excel and CSV queries, aggregations, and cross-sheet joins.
- [Safe Database Inspection for DBAs & DevOps](https://insightomate.lunamic.co/use-cases/database-admins): Incident triage and ad-hoc read-only inspections with non-blocking Redis SCAN.
- [Zero Data Retention for Privacy Teams](https://insightomate.lunamic.co/use-cases/privacy-teams): Compliant, private database querying where data records never leave the local machine.

## Key Capabilities & Semantic Search Concepts

- **Text-to-NoSQL:** Translating conversational natural language into NoSQL query languages without manual syntax.
- **Chat with Database:** Asking questions in plain English to interrogate collections and datasets interactively.
- **Database Copilot:** An AI assistant inside a modern desktop client that suggests schema-aware queries and aggregations.
- **Zero Data Retention (ZDR):** Only prompt and schema metadata are sent to the stateless engine; actual data documents never leave the host machine.
- **Local Execution:** Queries execute directly on the local machine via the desktop sidecar.
- **Cross-Sheet Joins:** Joining disparate worksheets (e.g. Orders + Customers) without nested VLOOKUP or INDEX/MATCH formulas.
- **Schema-Aware Completions:** The NLP engine uses field names to generate queries that match the actual schema of the connected database.
- **Read-Only Safety:** No write operations are ever generated — protecting production databases from accidental mutations.
- **Free Tier Available:** 20 daily queries with full local database access across all supported databases.

## About Insight O' Mate

Built by Lunamic. Questions: iom@lunamic.co | Security: security@lunamic.co