About Me
From Embedded Medical Devices to
Autonomous Agentic Ecosystems.
I am a multidisciplinary AI & Systems Engineer currently completing my Master's in Computer Science and Engineering at the University of Oulu (graduating June 2026). My work exists at the intersection of complex AI logic, resilient cloud infrastructure, and highly interactive user experiences. I don't just prompt language models , I build the distributed backend systems, vector databases, and multi-agent reasoning loops required to put them into production.
A Foundation in Logic & Interactive Systems
My journey didn't start in web development. I began in Electrical and Computer Engineering (B.Sc., University of Tehran), where I worked as an R&D Engineer developing algorithms for critical medical devices.
- 1
Interactive Systems & Game Development
In parallel, I discovered a deep passion for interactive systems — eventually co-founding a game development studio. I won the Best Bachelor's Project Award for EmoGalaxy 2, an interactive mobile game using on-device neural networks to help children with autism improve their emotion recognition skills. Building game loops taught me complex state management, performance optimization, and spatial UI — skills that now directly influence how I architect full-stack applications.
- 2
From Hardware Logic to Software Systems
Working at the intersection of embedded systems and interactive software gave me an intuition for performance constraints and deterministic logic that pure web developers rarely develop. Every millisecond matters whether you are running an IK solver on eight spider legs or running LLM inference on a constrained device.
Scaling Intelligence: Infrastructure & Agentic AI
As my projects grew in complexity, I realized that intelligent systems require bulletproof infrastructure. Today my focus is entirely on cloud-native backend engineering and Multi-Agent Systems.
- 1
Master's Thesis — UBICOMP Research Group, University of Oulu
I am architecting production-ready agentic workflows using the Microsoft Agent Framework, LangGraph, and PydanticAI — with full LLM observability instrumented via OpenLit for latency, token, and cost telemetry.
- 2
Production Systems at Scale
Whether engineering a high-performance RAG pipeline backed by Milvus, deploying a distributed SCADA system using Kafka and Kubernetes, or building a full-tenant SaaS ERP (Klaviro) with Next.js 15 and PostgreSQL — my goal is always the same: systems that are observable, scalable, and autonomous.
The Homelab & Continuous Learning
I am a firm believer in learning by deploying. Outside of academic and freelance work, I maintain a self-hosted micro-cloud PaaS environment using Dokploy and Docker Compose.
- 1
Cloudflare Zero Trust & Edge Monitoring
To ensure secure public ingress without exposing my local network, I route all traffic through Cloudflare Zero Trust Tunnels, with uptime monitored via Uptimeflare edge workers. This stack mirrors what enterprise teams run at significant cost — but entirely on my own hardware.
- 2
Experimental AI Sandbox
The homelab serves as my personal proving ground for testing bleeding-edge, sandboxed AI frameworks like Agent Zero and Ironclaw before adopting patterns into production work.
Technical Arsenal
The primary tools and technologies across the three pillars of my practice.
Agentic AI & Orchestration
LangGraph · AutoGen (Microsoft Agent Framework) · PydanticAI · OpenLit · LLM-as-a-Judge evaluation · Milvus · Qdrant · RAG architectures · Python
Cloud & Distributed Infrastructure
Docker · Kubernetes · Apache Kafka · MQTT · FastAPI · Dokploy · Cloudflare Zero Trust · MinIO · Nginx · TimescaleDB · PgBouncer
Spatial & Full-Stack Web
Next.js 15 · React · TypeScript · Node.js · PostgreSQL · Prisma · WebGL · Three.js · Unity · Unreal Engine 5 · C#
Let's build the next generation of intelligent systems.
I am actively seeking full-time roles starting Autumn 2026, with flexibility for part-time work sooner, in AI Engineering, Cloud Infrastructure, or Backend Development.