Workbench Overview
The full data platform: notebooks, dashboards, live queries, and AI-powered analysis.
What is the Workbench?
The Workbench is Thrive's full-featured data analytics platform, purpose-built for crypto traders and portfolio managers who need to go beyond pre-built dashboards. It combines the exploratory power of a Jupyter-style notebook environment with the operational polish of a modern BI tool like Looker, all connected directly to Thrive's real-time market data warehouse.
With the Workbench you can write SQL queries against live market data, run Python analysis scripts, build custom dashboards with drag-and-drop widgets, backtest trading strategies against historical data, and schedule reports to run on a recurring cadence. Every data source available on the Thrive platform, including derivatives, on-chain, sentiment, and trade history, is queryable from a single interface.
Pro+ and Founder exclusive
Core Capabilities
The Workbench is organized around three pillars: notebooks for exploration and analysis, dashboards for operational monitoring, and automation for recurring workflows. Each pillar is deeply integrated with the others, so a query you write in a notebook can be pinned to a dashboard and scheduled to refresh on a cron timer with no additional configuration.
Multi-cell notebooks with SQL, Python, AI, Markdown, visualization, backtest, and strategy cell types for end-to-end analysis workflows.
Drag-and-drop widget placement, resizable panels, template library, version history, and sharable public links.
Auto-refreshing queries at configurable intervals and streaming market data widgets for real-time monitoring.
Create calculated fields and composite metrics using a spreadsheet-style formula language across any data source.
Test trading strategies against historical market data with full P&L simulation and drawdown analysis.
Schedule any query or notebook to run on a cron timer. Results are cached and optionally emailed as reports.
Programmatic REST API for querying the Thrive data warehouse from external tools and scripts.
Subscribe to community-built signals and share your own analysis with other Thrive users.
The Data Warehouse
Every Workbench query runs against Thrive's PostgreSQL data warehouse, which aggregates data from all upstream sources into a unified, queryable schema. The warehouse is updated in near real-time with sub-minute latency for derivatives data and sub-hour latency for on-chain data. You have read-only access to every table in the schema.
Available Data Domains
| Market Data | OHLCV candles, aggregated prices, and market cap data across all tracked assets. Updated every minute. |
| Derivatives | Funding rates, open interest, liquidation events, and long/short ratios across Binance, Bybit, OKX, and Hyperliquid. |
| On-Chain | Exchange netflows, active addresses, large transactions, smart money holdings, and DEX volume. Nansen-powered. |
| Signals & Divergences | Historical AI signals with confidence scores and outcomes, plus all detected divergences with resolution data. |
| Trade History | Your personal trade journal data including entries, exits, P&L, tags, and notes. Only your own data is accessible. |
| Social Sentiment | Aggregated sentiment scores, trending narratives, KOL activity, and social volume by asset. |
| Screener Snapshots | Point-in-time snapshots of the asset screener, enabling historical analysis of z-scores and rankings. |
| Calendar Events | Token unlocks, protocol upgrades, exchange listings, and economic events with impact scores. |
Schema Explorer
The Workbench includes a built-in schema explorer accessible from the sidebar of any notebook or dashboard. It lists every available table, column names and types, sample data, and relationship diagrams. Click any column name to auto-insert it into the active SQL cell. The schema explorer also supports full-text search so you can find the right table by searching for a keyword like "funding" or "liquidation."
Getting Started
The fastest way to start using the Workbench is to open a pre-built template notebook and run its cells. This gives you a feel for the query workflow before you write anything from scratch.
Navigate to the Workbench
Click Workbench in the sidebar or press Cmd+K and type "workbench." You land on the Workbench home page showing your recent notebooks, saved dashboards, and template library.
Open a template notebook
Click Templates and select one of the pre-built notebooks. "Market Overview" is a good starting point: it queries current funding rates, z-scores, and open interest across the top 20 assets and renders the results in a sortable table and heatmap.
Run all cells
Click Run All in the notebook toolbar or press Shift+Enter on each cell sequentially. SQL cells execute against the data warehouse and display results inline. Visualization cells render charts from the query output.
Modify and experiment
Edit the SQL in any cell to change the query. For example, add aWHERE clause to filter for a specific asset, or change the timeframe. Re-run the cell to see updated results. This iterative workflow is the core of the notebook experience.
Save your work
Click Save or press Cmd+S to persist the notebook. Notebooks are auto-saved every 60 seconds, but manual saves create named versions you can revert to later.
Workbench vs. Other Tools
The Workbench replaces several standalone tools in a typical crypto analyst's stack. Understanding these comparisons helps clarify what the Workbench can do for you.
| Jupyter Notebooks | Similar multi-cell paradigm with SQL and Python, but pre-connected to live crypto data. No environment setup, no dependency management, no data pipeline configuration. |
| Looker / Metabase | Similar dashboard builder and scheduled reports, but with notebook-level query flexibility and crypto-specific data models. |
| Dune Analytics | Similar SQL-on-blockchain-data concept, but Thrive includes derivatives, CEX data, and trade history alongside on-chain data. No Solidity or EVM expertise required. |
| Google Sheets | The formula builder provides spreadsheet-like calculated fields, but connected to live market data rather than static imports. |
| TradingView Pine Script | The backtesting engine lets you test strategies similarly to Pine Script, but with access to derivatives, on-chain, and sentiment data that Pine Script cannot access. |
Credit Costs
Workbench operations consume credits based on the computational resources they require. Simple SQL queries are inexpensive; Python execution and AI-powered analysis cost more.
| SQL query | 5 credits per execution. Covers data warehouse queries of any complexity. |
| AI cell (natural language to SQL) | 15 credits. The AI generates and executes the query on your behalf. |
| Python cell execution | 20 credits. Runs in a sandboxed environment with pre-installed data science libraries. |
| Backtest (simple) | 25 credits. Single-asset strategy with up to 90 days of history. |
| Backtest (complex) | 50 credits. Multi-asset or full-year history with advanced parameters. |
| Formula evaluation | 2 credits per formula per refresh cycle. |
| Dashboard refresh | Sum of all widget query costs. A dashboard with 6 SQL widgets costs 30 credits per full refresh. |
| Scheduled query run | Same cost as manual execution. Deducted from your monthly allocation at run time. |
Optimize with caching
Next Steps
Notebooks & Cells
Master every cell type: SQL, Python, AI, visualization, backtest, and more.
Dashboard Builder
Build operational dashboards with drag-and-drop widgets and templates.
Live Queries & Streaming
Set up auto-refreshing queries and real-time data feeds.
Advanced Workbench
Formula builder, backtesting, scheduling, and the signal marketplace.