Blabber.ai

Blabber.ai is a creator-focused communication platform that facilitates voice and messaging automation.

• Subscription sync failures between Stripe and Supabase resulting in access resolution issues.
• Webhook queue delays under concurrent payment events.
• Security vulnerabilities in access control checks on public API endpoints.
• High Sentry exception volumes prior to public release.

• Zero-Downtime Hardening: System upgrades had to be deployed to the live staging environment without interrupting ongoing user testing.
• API Latency Constraints: Supabase database latency and external webhook handlers had to respond within a 500ms SLA.
• Strict Compliance: Payment and session management flows had to adhere to security best practices without introducing heavy operational overhead.

SazM executed a launch readiness audit, profiling database queries and API routing structures. A decoupled webhook handler was introduced to process Stripe subscription lifecycles asynchronously, and access controls were hardened at the Supabase Row Level Security (RLS) and API gateway levels.

• Frontend & Routing: React and Next.js frontend integrated with Next.js API routing.
• Database & Auth: PostgreSQL database hosted on Supabase, secured with custom Row Level Security (RLS) policies and session-based authentication checks.
• Payment & Event Tier: Decoupled Stripe webhook receiver with retry and idempotency validation layers.

• Production readiness audit and launch validation checklist.
• Hardened Supabase database schema with optimized PostgreSQL queries.
• Idempotent Stripe webhook handler with built-in retry mechanisms.
• Decoupled API endpoint security validations.

• Webhook integrations must be idempotent at the database layer; network retries will duplicate incoming events.
• Client-side security checks are insufficient; database-level security policies (like Postgres RLS) are critical to defend multi-tenant architectures.