#160: Implement real API for KPI Analytics Lambda

[anupbpote02]

Summary

Implements kpi_api_analytics.py — a new AWS Lambda that connects to the Virginia RDS PostgreSQL database and returns chart-ready JSON for two KPI dashboard widgets.

Closes #160

What's Implemented

Part 1 — Request Status Distribution

• Counts requests grouped by status (CREATED, IN_PROGRESS, RESOLVED)
• Returns request_status_distribution array and total_requests count
• Returns [] and 0 when no data exists

Part 2 — Average Resolution Time by Category

• Calculates avg resolution time (hours) per help category
• Uses submission_date and serviced_date, only includes fully resolved requests
• Returns SLA metadata (target: 10 days/240 hours, warning: 8.33 days/200 hours)
• Returns [] when no resolved requests exist, SLA always present

Error Handling

• DB connection and queries wrapped in try/except
• Returns statusCode 500 with safe defaults on failure
• finally block closes cursor and connection

Local Testing

• __main__ block for local execution
• Verified: safe response on missing env vars, no crashes, all keys present

Schema Corrections from Issue

The issue referenced table/column names that differ from the actual Virginia DB:

Issue assumed	Actual DB	Used in code
request.request_id	request.req_id	req_id
request.request_status	request.req_status_id (FK)	Joined to request_status table
request.category_id	request.req_cat_id	req_cat_id
request_category table	help_categories table	help_categories
request.resolved_date	request.serviced_date	serviced_date

Reference

• Follows patterns from beneficiariesTrendAnalysis.py and volunteer_application_analytics.py
• Table/column names confirmed from Saayam_Table.column.names_data.xlsx