What this PR does / why we need it:

Adds a blog post extending Feast Feature Server Monitoring to cover two new observability capabilities introduced in PR #6340:

1. Offline Store RED Metrics — Prometheus counters and histograms (feast_offline_store_request_total, feast_offline_store_request_latency_seconds, feast_offline_store_row_count) for offline store retrieval operations, closing the observability gap between online serving and offline training paths.

2. SOX Audit Logging — Structured JSON audit log entries emitted via a dedicated feast.audit Python logger for both online and offline feature access, designed for compliance with SOX, GDPR, HIPAA, and similar regulatory frameworks.

The blog covers:

• Why offline store metrics and audit logging matter in production ML systems
• The three new Prometheus metrics and their label structure
• SOX audit log entry format for online and offline paths (with PII-minimization by design)
• YAML configuration to enable offline_features and audit_logging
• PromQL queries, Prometheus alert rules, and extended Grafana dashboard panels
• How to route audit logs to compliance sinks (Splunk, ELK, S3 WORM)
• Step-by-step demo using the feast-prometheus-metrics setup

Which issue(s) this PR fixes:

Documentation for PR #6340 (Offline store metrics and SOX audit logging)

Checks

• I've made sure the tests are passing.
• My commits are signed off (git commit -s)
• My PR title follows conventional commits format

Testing Strategy

• Testing is not required for this change (documentation only)