Building Production AI Applications
About Me
Muhammad Ahsaan Ullah is an AI Systems Engineer focused on building production-ready AI applications, agentic systems, and scalable backend platforms. His work spans AI automation, FastAPI, multi-tenant SaaS, computer vision, and retrieval-augmented generation (RAG). He has built systems including TarkaBot, a WhatsApp Restaurant Operating System, DeepFakeShield, an AI-powered media forensics platform, and ExportShield Pro, a ZATCA compliance solution. He enjoys solving real business problems by combining machine learning with robust software engineering.
Unlike traditional software engineers, I approach systems with a strict product mindset. I do not just write code; I architect systems that reduce manual work, streamline operations, and directly impact the bottom line. From processing massive 350K+ face datasets for DeepFake detection to building multi-tenant SaaS architectures, my goal is always to deliver production-grade systems.
Validation & Milestones
DeepFakeShield Platform
Built a production-grade deepfake detection system achieving 95.83% validation accuracy across a 3-class classification pipeline.
TarkaBot AI
Successfully architected and deployed a multi-tenant WhatsApp automation SaaS designed for restaurant operations.
ExportShield Pro
Built Fatoorah-compliant e-invoicing middleware meeting strict cryptographic requirements by the Saudi ZATCA authority.
