I build AI systems that actually run in production.
RAG pipelines, multi-agent automation, and SaaS platforms — architected, coded, deployed, and monitored. End-to-end, no handoffs. Self-hosted where it makes sense. Cloud where it doesn't.
About
Most AI projects die between demo and production. I close that gap — database schema, backend API, dashboard UI, authentication, CI/CD, monitoring, and security hardening, all in one engagement.
I specialize in production RAG systems, multi-agent workflow automation, and SaaS multi-tenancy. Everything runs on self-hosted ARM64 infrastructure — Docker, Nginx Proxy Manager, Jenkins CI/CD, and real-time monitoring.
What I Deliver
AI Engineering
Production RAG — Document Q&A pipelines that stay accurate at scale. Self-hosted embedding, sub-300ms query on 9K+ docs.
Multi-Agent Orchestration — Complex tasks into parallel sub-agents. Token-efficient with Beads task graph, real-time dashboard.
SaaS Platform
Multi-Tenant SaaS — Isolated data per tenant, multi-level RBAC, QRIS payment gateway, automated testing, zero-downtime deployment.
Self-Hosted Infrastructure
ARM64 server — Docker orchestration, Nginx Proxy, CI/CD pipeline, monitoring, 14 CVEs patched without downtime.
Selected Work
SIMasjid / As-Shafa
Masjid management system — member database, donation tracking, event scheduling. Laravel + Convex + self-hosted.
HRIS ITB
HR information system for Institut Teknologi Bandung — employee management, leave, payroll integration.
OpenWA CLI
WhatsApp API gateway CLI — send/receive messages, manage sessions, webhook integration.
Agent API Gateway
Multi-model AI endpoint — one API, 10+ providers, auto-routing, cost logging, audit trail.