Platform & Tools

All data sources are accessible through a single RESTful API with consistent schemas, standardized units, and unified geographic identifiers. Query emissions, air quality, and inventory data through one interface, with cross-source joins built in: air-quality stations linked to nearby emission facilities, EDGAR grid cells reconciled with Climate TRACE assets, all in a single query. Filtering, pagination, aggregation, and ESG-specific endpoints are first-class. OpenAPI documentation. One auth, one token, one set of rate limits.

For developers, data engineers, platform integrators, and ESG analysts.

Jana's native Python client library handles authentication, pagination, and every endpoint with one import. Works in Jupyter Notebooks and Google Colab. Device-code login means no keys to copy-paste. Pre-built example notebooks for emissions analysis, air-quality spatial joins, and historical trend studies let you go from blank notebook to working analysis in minutes.

For data scientists, researchers, and analysts who live in Python.

Jana's geospatial data works natively with QGIS, ArcGIS, Google Earth Engine, and any standard GIS workflow. EDGAR's 0.1-degree gridded layers, Climate TRACE facility coordinates, and OpenAQ station locations import cleanly. Visualize emissions on maps. Overlay air quality with industrial activity. Run buffer-zone analyses around emitting sites. Build population-exposure models. Produce publication-quality cartography from the same data your colleagues are querying programmatically.

For urban planners, environmental consultants, policy analysts, journalists, and anyone who works with maps and spatial analysis.

A Model Context Protocol server connects AI tools (Claude, Cursor, ChatGPT) directly to Jana's live data. Your AI assistant queries real measurements instead of training-data approximations. 

Use natural language to explore datasets, detect trends, identify anomalies, and run cross-source comparisons, all without leaving your chat or IDE.

Every AI-generated answer surfaces the data it pulled, the methodology of each source, and the limits of the inference. No black boxes. Provenance is preserved through the model, not lost at the prompt boundary, so results are suitable for academic publication and audit.

For researchers, analysts, and anyone who wants AI acceleration without losing defensibility.