• Independently own the modernization of Magiran, one of Iran's largest academic indexing platforms, with end-to-end technical ownership spanning architecture, database design, implementation, deployment, and long-term operations.
• Led the modernization of a large-scale legacy platform by redesigning core data storage and caching strategies, significantly improving application performance, reducing I/O overhead, and enabling the platform to efficiently serve hundreds of thousands of daily visitors and millions of requests.
• Rebuilt the platform's search infrastructure and query processing capabilities, delivering a 5–10x improvement in search performance across a Persian-language index containing more than five million records.
• Designed and implemented a highly automated content-ingestion pipeline for magazine publications, overcoming complex and heterogeneous XML formats from multiple sources and reducing publication time from 10–20 minutes to a few seconds (a 20–50x improvement in throughput).
• Redesigned core database structures and executed large-scale data migrations supporting a production database exceeding 412 GB while improving scalability and maintainability.
• Designed and maintained production infrastructure, deployment pipelines, monitoring, and administration systems to ensure platform reliability and support editorial and operational workflows.
• Advocated for additional engineering headcount to support growing scope; when unavailable, maintained full technical ownership from architecture through delivery.

Independently designed, developed, and operated

• Architected the platform using Clean Architecture and Domain-Driven Design, with CQRS via MediatR to separate business logic from infrastructure concerns.
• Implemented optimistic concurrency control and row-level pessimistic locking to guarantee payment and order consistency under concurrent load.
• Built and containerized the full production stack — .NET 10 Minimal API backend, Next.js/MUI frontend, PostgreSQL, Redis for caching, Nginx, and centralized logging — deployed via Docker Compose; self-hosted initially and now migrating to a dedicated VPS for production scale.
• Developed a Persian-language full-text search engine with a custom analyzer for accurate text matching.
• Designed a fault-tolerant background job processing system with distributed locking and graceful cancellation for asynchronous order and notification handling.
• Designed and implemented zero-downtime full-text search infrastructure optimized for Persian-language content.