DOCSEARCHbyAlgolia
Submit Docs

Supersimple Documentation

website·Indexed 33 days ago

https://docs.supersimple.io

A data exploration platform featuring no-code exploration steps, AI-powered analysis, and flexible data modeling. The platform supports SQL, Python, and TypeScript code blocks, custom formulas, and integrations with Slack, GitHub, Confluence, and the Model Context Protocol (MCP). It includes tools for data visualization, model access control, and a CLI for managing data models and metrics.

Tokens
28.4K
Snippets
133
Records
243
Agent score
57%

What's inside Supersimple

  1. Overview of Supersimple BI Platform

    Supersimple is a self-service business intelligence platform designed to allow team members to answer complex data questions without relying on static dashboards. It operates by connecting directly to a user's data warehouse and unstructured context sources (such as Slack, Google Drive, GitHub, and Confluence) to provide AI-driven insights. Crucially, Supersimple does not store the user's data; it fetches fresh data directly from the data warehouse whenever a report is opened or a question is asked.

  2. Use fundamental exploration step types

    Supersimple provides four fundamental step types to manipulate data models:

    • Filter: Removes rows based on specific criteria.
    • Summarize: Groups data and performs calculations.
    • Jump to Related Data: Navigates to related data models for deeper analysis.
    • New Column: Adds new columns using custom formulas or data model relations.

  3. Understand Supersimple data models

    Supersimple's data modeling layer allows you to define business-level concepts (e.g., Organizations, Contacts, Deals) that internal users and AI use to explore data. It acts as a semantic layer that maps raw data warehouse tables to a structured mental model of the business.

    For every business concept, the data model defines:

    1. Source: The specific table in the data warehouse where the data originates (e.g., deal_activity_email).
    2. Fields: The properties or columns associated with that model (e.g., a Company's name).
    3. Relations: The logical links between different data models (e.g., linking a Ride to a Driver).

    Developers can use the data model to restructure or clean data centrally, such as filtering out test records (e.g., where is_test is true) or merging multiple raw tables (e.g., combining emails_sent_via_ui and gmail_sync_emails) into a single unified business concept.

  4. Reuse logic using Instance blocks

    Instance blocks allow you to reuse logic across an exploration by linking a new block to a 'master' block. When a block is duplicated as an instance, any changes, filters, or steps added to the master block are automatically carried over to all its instances. Master blocks are identified by a specific icon on the block and in the sidebar, which also indicates where the master is being used.

  5. Configure AI Instructions for persistent business context

    Instructions provide the AI with durable, specific, and actionable context about your company, terminology, and data. The AI applies these based on the scope of the query:

    • General instructions: Apply to every connected data source (e.g., company-wide terminology, KPIs).
    • Database connection instructions: Apply when the AI works with models/data from a specific connection.
    • Integration instructions: Apply when the AI searches content from integrations like Slack, Notion, or Google Drive.

    Steps to configure:

    1. Go to Account Settings → AI → Instructions.
    2. Select the desired scope (General, Database, or Integration).
    3. Enter the context and select Save.

  6. Use Slack as a data source for enterprise search

    Connecting Slack allows Supersimple's enterprise search and data questioning tools to include conversation context. This includes relevant threads from internal channels and Slack Connect channels shared with external parties, providing context for customer issues, technical debugging, or the reasoning behind specific metrics and data models.

  7. Use Supersimple for Customer Success workflows

    Customer Success teams can utilize Supersimple to monitor customer health and drive growth through the following capabilities:

    • Behavioral Notifications: Get alerted to critical customer behaviors, such as churn risks or potential expansion opportunities.
    • Customer Health Checks: Quickly retrieve a comprehensive overview of specific customer accounts.
    • Onboarding Optimization: Guide customers toward their 'aha' moment during the onboarding process.
    • Contextual Support: Access necessary customer context to resolve specific pain points and struggles.
    • Goal Tracking: Monitor and track progress toward defined customer goals.

  8. Query data using Supersimple AI natural language exploration

    Supersimple allows users to ask questions about their data using natural language. When a question is submitted, the AI automatically constructs a mini-report by assembling the same no-code exploration steps (like Filter, New Column, or Summarize) that a user would manually select.

  9. Understand Supersimple AI's grounding and data access

    Supersimple's AI is grounded in a governed semantic layer, ensuring that AI-generated work uses the same models, relationships, metrics, and permissions as human-created work. Additionally, the AI respects data-access configurations; teams can decide whether the AI is permitted to inspect query results.

  10. Understand the Supersimple no-code exploration pipeline

    Supersimple uses a sequential pipeline of exploration steps to process data. Every exploration block begins with a selected data model, which initially displays all rows and columns from that model. Users then add steps via the sidebar to refine the data. These steps are applied top-to-bottom, meaning each step operates on the output of the previous one.

  11. Connect a data warehouse via OpenVPN

    Supersimple can connect to databases or data warehouses that are not exposed to the internet by running a persistent OpenVPN client. This creates a long-lived tunnel where only traffic to the specific database address and port is routed. Supersimple only initiates outbound connections to the VPN endpoint and the database; no inbound connections to Supersimple are required.

  12. Understand and use the Supersimple Inbox for data model improvements

    The Supersimple Inbox is a focused queue of improvements for your data setup. After an AI question is answered, Supersimple reviews the agent's work in the background (typically within 10–20 seconds) to identify structural gaps that hindered accuracy or efficiency.

    Common gaps identified include:

    • Missing data models, properties, metrics, or relations
    • Undefined or ambiguous business terms
    • Categorical values missing enum documentation
    • Business context residing in external sources (e.g., Slack, Notion)
    • Missing, stale, or inconsistent data

    Access Control: Inbox is visible to account admins and users with the 'Manage Account Settings' permission. It is accessed via the main sidebar.