AGENTIC AI PROJECTS
Alwakala AI.
Brand-aware content workflows, from planning and generation to quality checks and publishing.
Full-stack development & AI workflows
OGz Studios · Multi-agent content platform
Next.js · TypeScript · n8n · Supabase · RAG · LangChain
The problem
Producing consistent content for multiple brands means coordinating brand context, planning, writing, and review across many separate steps.
The solution
A RAG-based multi-agent platform connects brand onboarding, content planning, generation, quality checks, and publishing. Alwakala is the live product associated with the OGz Studios project.
What I built
- Orchestrated 11 n8n workflows and 18 content chains using GPT-4o and Claude.
- Connected brand context with retrieval using Supabase and Qdrant.
- Added schema validation for model responses before they move through the workflow.
What it brings together
Reduced nightly content generation for approximately 60 brands from about 18 hours to about 2 hours, as reported in my resume. Schema validation keeps model outputs structured before they move through downstream steps.