I need a production-ready REST API that returns block-level real-estate intelligence for every street segment in Montreal. The service must synthesise raw public records and third-party feeds into three core metrics—price trends, rental yield, and an overall investment score—so real-estate investors can instantly compare one side of the street to the other. Scope of work You will design the full data pipeline: ingest (scraping or bulk downloads), clean, geocode to individual blocks, run the analytics model, then expose the results through well-documented JSON endpoints. Each request should accept latitude/longitude or standard address strings and respond in under 300 ms. Key expectations • Daily or near-real-time refresh of source data • Clear explanation of the price-trend and yield calculations so figures can be audited • Geo-aggregations no larger than a single city block (≈50–100 m) • Dockerised deployment plus a small demo script that queries the API and prints the metrics Acceptance criteria • Endpoint /v1/block-insight returns current price trend %, rental yield %, and investment score for a valid coordinate and HTTP 200 with <300 ms latency on a small VPS • Model accuracy: ±5 % when back-tested against last year’s closed sales • Complete README covering setup, dependencies, and example cURL requests Preferred stack is Python (Pandas, GeoPandas, PostGIS, FastAPI) or Node.js with equivalent geospatial tooling, but I’m open if you can demonstrate similar performance and maintainability. Let me know how you plan to source the data, outline your modelling approach, and share a link to a previous geospatial or fintech API you’ve built.