Mahdy Muzzammil

Building scalable backends and agentic workflows

---

🛠️ Technical Skills

• Languages & Frameworks: TypeScript, Python, Node.js (Express), FastAPI, LangGraph, LangSmith, JavaScript
• Cloud & Infrastructure: AWS (EC2, ECS, EKS, ELB, S3, SQS, SNS), Docker, BullMQ
• Databases & Caching: PostgreSQL, MySQL, DynamoDB, Redis (Pub/Sub, Streams), Prisma ORM

🚀 Featured Projects

🔴 Backend Systems

api_monitor — TypeScript Production-grade API monitoring & incident tracking system Queue-first async processing with BullMQ + Redis workers Rolling analytics: 24h/7d/30d uptime, avg latency, p95/p99 Incident workflows with OPEN/ACKNOWLEDGED states Webhook alerts with retry/backoff + full Swagger/OpenAPI docs

prokoi — Python Multi-tenant project management backend for organizations API → Service → Repository architecture with FastAPI JWT auth, custom middleware, permission-based RBAC Sprint lifecycle APIs: velocity, workload, team performance Redis Streams + WebSocket real-time notifications

🤖 AI & Agentic Systems

TA-AI — Python + TypeScript Automated evaluation & curriculum-aware teaching assistant Multi-agent LangGraph pipeline: Vision LLM extracts questions, rubrics, solutions, and student answers from uploaded PDFs via a shared graph-builder factory Dual-grader ensemble with strict + fair LLM scoring, confidence aggregation, and per-question iterative persistence GraphRAG syllabus agent combining vector search + entity-relationship graph to power a TA chatbot with curriculum-aware context

EpiTrace — JavaScript Autonomous incident monitoring & remediation backend Designed for "Cline CLI as Infrastructure" workflows Detects, traces, and remediates backend incidents Agentic approach to ops — no manual intervention

---