
Saad Rupai
Backend Architect for AI, SaaS and Scalable Systems
Competenze

Consulta i miei servizi


Portfolio
Esperienza lavorativa
Software Engineer - Payments & AI Agent Pipelines
Shikho
May 2026 - Present • 6 mos
Backend engineer at one of Bangladesh's largest EdTech platforms (2.5M+ active users). - Designed and built an AI Agent pipeline in Go that turns lecture videos and textbooks into personalized quizzes, model tests and chapter summaries. - Built and scaled the payment processing backend handling transactions for 2.5M+ active users, integrating multiple payment gateways with high availability. - Integrated campaign features with Grameenphone (MyGP) for seamless cross-platform authentication and content delivery. - Led database optimization and schema migrations on large datasets and built real-time monitoring dashboards.
Software Engineer (Remote) - Billing, Metering & AI Streaming
Action Board AI
Dec 2025 - Aug 2026 • 8 mos
Remote backend engineer for Actionboard.ai (by Cloudscockpit, Maryland, USA), an enterprise AI workflow automation and subscription billing platform. - Engineered a scalable billing and usage-metering platform in Go supporting subscription- and usage-based pricing, proration and automated Stripe / SSLCommerz payments. - Architected an asynchronous ML response streaming pipeline with AWS SQS, AWS Lambda (response streaming Function URLs) and Redis Streams, delivering chunks to UI clients with sub-100ms latency. - Integrated OpenRouter LLMs and GPU compute nodes into the streaming pipeline. - Built an event-driven notification service with RabbitMQ and Redis for real-time alerts across web and mobile.
Software Engineer - Backend Team Lead
TechnoNext
Apr 2025 - May 2026 • 1 yr 1 mo
Backend team lead at the technology arm of the US-Bangla group. - Led the backend team building a distributed order-processing platform handling 10,000+ order requests at peak, using an asynchronous message-broker architecture and RPC communication. - Migrated 600,000+ geospatial records to a search engine, achieving sub-100ms p99 latency for geocoding and reverse-geocoding. - Built a high-performance geo-service powering real-time location intelligence across multiple internal platforms. - Decoupled order processing from a monolithic inventory system into a microservice, cutting average response time by ~32% under peak load.