Moamen Osama

I am a passionate AI Engineer. I love my field and it's my minor in college. My goal is to be able to contribute with the skillset I have in real live situations and help shaping the future.

AI Engineer at RockAI Dev (2024-01 – Present)

CRM AI Assistant, FastAPI, LangChain, LangSmith, MongoDB, uv

• Engineered a multi-tenant chatbot for sales teams featuring Role-Based Access Control (RBAC) and dynamic identity-aware system prompt injection at runtime.
• Optimized token efficiency by designing a Router LLM to selectively inject tool schemas per request, reducing prompt overhead and improving end-to-end response latency.
• Implemented nested tool execution with manual piping to prevent context-window bloating caused by chained outputs in complex multi-step workflows.
• Established robust observability via LangSmith; developed intent-routing evals and implemented real-time prompt injection detection with trace-visible attempt counters.
• Architected a property-discovery system using Qdrant with advanced payload filtering for complex structured queries including budget, location, and property attributes.
• Integrated a Dual-LLM pattern to extract structured filters, ensuring vector lookups are only triggered for relevant intents to preserve compute and reduce costs.
• Built a behavior-based personalization layer that tracks user search history to surface high-relevance recommendations grounded in real estate financial modeling.

Bachelor of Science in Data Science, Minor in Artificial Intelligence – Faculty of Computers and Data Science – Alexandria University (2021-11 – 2026-06)