Connects to your stack
SnowflakeBigQueryPostgres
RedshiftClickHouseDatabricksDuckDB
RedshiftClickHouseDatabricksDuckDBA governed analytics platform that turns natural-language questions into accurate SQL against your own data. Built on a Rust query engine, a semantic layer for consistent metrics, and your choice of LLM provider — deployed on infrastructure you control.
RedshiftClickHouseDatabricksDuckDBConversational querying is just the surface. Underneath sits a governed semantic layer, a Rust execution engine, and the access controls a real production platform needs.
Conversational analytics for the whole team. Ask follow-up questions, refine filters, and inspect the generated SQL before it runs. No prompt engineering required.
Define metrics, dimensions, and relationships once. The Modeling Definition Language guarantees every answer uses the same governed definitions across teams.
Snowflake, BigQuery, Redshift, Postgres, MySQL, ClickHouse, DuckDB, Trino, Athena, Databricks and more. Bring your own warehouse — no data movement, no extracts.
Every answer renders as an interactive chart. Pin questions to dashboards, schedule refreshes, and share read-only or editable links with your team.
Beyond the chart: surface anomalies, year-over-year deltas, segment outliers, and likely root causes — written in business language, not SQL.
Admin, Editor, and Viewer roles built in. Email/password login or Google and GitHub OAuth. Dashboards and conversation threads can be shared view-only or editable, per user.
OpenAI, Anthropic, Google, Azure, Databricks, DeepSeek, Groq, or fully local Ollama. Switch providers without re-modeling your semantic layer.
Semantic models live as YAML in your repo. Diffs, reviews, and CI checks like any other code. Roll back a bad metric definition in seconds.
Starred dashboards, shareable conversation threads, comments on saved questions. Designed for analysts, PMs, and ops to work side-by-side.
MDL — our Modeling Definition Language — defines tables, columns, relationships, and business metrics once. Every natural-language question routes through it, so an answer to "what's MRR?" is the same whether it comes from sales, finance, or a Slack thread.
SUM(subscriptions.mrr) WHERE cancelled_at IS NULLCOUNT(*) FILTER (status = 'churned') / COUNT(*)Pick your warehouse, pick your model provider. We never copy your data — queries execute where your tables live.
11+ first-class connectors today, with an extensible driver SDK for in-house systems.
Redshift
Azure SynapseDatabricks
SQL Server
OracleDuckDBClickHouseTrinoAthenaUse a hosted frontier model, your own cloud's foundation model, or a fully local model — your choice, swappable per workspace.
Azure OpenAIAWS BedrockDatabricks
DeepSeek
GroqOllama (local)Every step is observable, governed, and replaceable. No black-box magic — just five clearly named services you can audit and operate.
A user asks a question in natural language from the UI, Slack, or the API.
Intent classification, schema retrieval from Qdrant (RAG), and SQL generation via your chosen LLM.
Generated SQL is rewritten through the MDL to use governed metrics, joins, and access rules.
The Rust analytics engine plans and executes against your warehouse — no data leaves your VPC.
Results come back as a chart, table, and natural-language summary, ready to pin to a dashboard.
NQRust-Analytics is sold as a single enterprise tier with two ways to run it: installed on your own infrastructure, or hosted by our team in a dedicated single-tenant cloud. We don't run a public, multi-tenant SaaS — your warehouse shouldn't share a database with anyone else's.
The full NQRust-Analytics platform, deployed on your infrastructure or managed for you in a dedicated single-tenant cloud — installed, configured, and supported by our team.
Self-hosted — install on your VPC, bare metal, or private cloud. You control the network, the LLM provider, and the upgrade cadence.
Managed single-tenant — we run a dedicated instance for you on a cloud of your choice. Same software, no shared database, no shared compute.
Deployment scope, support level, and pricing are tailored per customer. Contracts are annual; we'll walk you through what a typical engagement looks like.
We're a small team — every inquiry goes straight to a human, usually within one business day. Tell us about your warehouse and what you're trying to answer, and we'll get back to you with next steps.
Your data warehouse, team size, preferred deployment model (self-hosted or managed single-tenant), and the questions your team is currently struggling to answer.
NQRust-Analytics is a commercial, closed-source platform. Reach out for an evaluation, a guided demo with your own schema, or a quote sized to your deployment.