Tech Stack

Technology Stack & Prototyping Strategy

To ensure maximum agility, we are initiating development with a local-first prototyping strategy. This approach bypasses an immediate AWS deployment for two primary reasons: first, it significantly accelerates the development cycle by eliminating cloud overhead during the initial build. Second, it allows us to make meaningful progress while client-side AWS account provisioning is finalized.

Component	Phase 1: Local Development	Phase 2: AWS Cloud (Production)
Scraper / Ingestion	Python (Requests / BeautifulSoup)	AWS Lambda / Managed Scraping APIs
Data Lake	Local File Storage	Amazon S3
Database	DuckDB	Amazon Athena
Data Format	Parquet / JSON	Parquet / JSON
Automation / Trigger	Manual / GitHub Actions	AWS EventBridge
UI / Dashboard	Streamlit (Folium / Pydeck)	AWS App Runner (or React/FastAPI)
CI / CD	GitHub Actions (Test & Lint)	GitHub Actions (Automated Deployment)
Infrastructure	Manual Configuration	AWS SAM (Infrastructure as Code)
Intelligence / AI	scikit-learn / Pandas	AWS Lambda (Scikit-learn Layers)