Geospatial Analysis Website & API

Заказчик: AI | Опубликовано: 04.04.2026

I’m building a web platform focused on geospatial data analysis and I want it backed by a clean, well-documented REST API for mobile device communication. The core of the project is user define area: users will query our private spatial datasets, run server-side calculations (buffering, distance, clustering, custom stats), and instantly see the results rendered on screen. Maps The front end should support both interactive and static map views along with user entered data. When users zoom, pan, or apply a filter they should be able to select a collection of cells which are used to create a user define area/zone, it should fall back to a high-resolution static image. Heat-map or other thematic layers could be added later, so keep the architecture flexible. Data sources All data comes from our private repositories—PostGIS tables and a secure object store. No public feeds are needed today, but the API should be modular enough to accept additional sources later. Key expectations • Responsive web UI (React, Vue, or similar) using Leaflet, Mapbox GL JS, or another modern mapping library. • RESTful back-end (Python/FastAPI, Node/Express, or comparable) with endpoints for authentication, spatial queries, and results export (GeoJSON, CSV, PNG). • Spatial processing handled server-side—prefer PostGIS or a lightweight GeoPandas workflow. • Clear API documentation (OpenAPI / Swagger) and a short README so my internal team can extend it. • Deployment scripts or Docker files that run end-to-end on a standard cloud VM. Acceptance criteria 1. A user can sign in, select cells for their user-define area/zone. Modify parameters and other fields. 2. The same request, when called via cURL, returns identical GeoJSON. 3. Static export produces a 300 dpi PNG matching the on-screen view. 4. All endpoints are covered by unit tests (≥80 % coverage) and documented in Swagger UI. See attach for more information. If parts of the stack are your specialty—say you excel at Mapbox or PostGIS—let me know; I’m open to recommendations as long as the final product meets the criteria above and keeps performance tight.