const parsa: AIEngineer = {
name: "Parsa Shadkam",
role: "Senior Backend → AI Software Engineer",
location: "Tehran, Iran 🇮🇷 (Open to Remote 🌍)",
education: "B.Sc. Computer Engineering @ KNTU",
specialties: [
"Scalable Backend Systems",
"AI / ML Integration & Infra",
"Microservices & Cloud-Native",
"Clean Architecture (DDD, Hexagonal)"
],
stack: {
languages: ["TypeScript", "Python", "C++"],
backend: ["NestJS", "Django", "FastAPI"],
ai: ["PyTorch", "TensorFlow", "LangChain"],
cloud: ["Docker", "Kubernetes", "AWS"]
},
currentFocus: "Bridging production-grade backends " +
"with intelligent AI systems",
funFact: "I treat every system design like an ICPC " +
"problem — optimize first, scale forever.",
motto: "Code is a hypothesis. Ship the experiment. 🚀"
};
✨ A curated tour through my codebase — from production microservices to algorithm training grounds ✨
|
A full-fledged e-commerce platform built on microservices architecture using NestJS and Turborepo monorepo tooling. Features independently deployable services, Docker-based orchestration, and pnpm workspaces for scalable backend + frontend coordination. 💡 Demonstrates: service decomposition, monorepo design, container orchestration |
A reference implementation of Hexagonal Architecture (Ports & Adapters) in NestJS — separating domain logic from infrastructure. Showcases Domain-Driven Design, dependency inversion, and a clean testable core independent of frameworks. 💡 Demonstrates: DDD, clean architecture, framework-agnostic domain modeling |
|
REST API powering StoreTrack, an inventory & store management platform. Handles product catalogs, stock tracking, and order workflows with NestJS + PostgreSQL + TypeORM. 💡 Demonstrates: REST API design, relational modeling, business logic encapsulation |
Modern, responsive React interface for StoreTrack — with protected routes, reactive forms, and seamless integration with the NestJS backend. Paired companion to the StoreTrack API. 💡 Demonstrates: full-stack integration, component-driven UI, state management |
|
A production-ready task management REST API featuring JWT authentication, role-based access, CRUD workflows, and database migrations. Built with clean, modular NestJS patterns. 💡 Demonstrates: auth flows, modular architecture, CRUD best practices |
🎯 JustTact
A focused backend project exploring tactical system design and clean modular patterns. A personal sandbox for experimenting with architectural ideas in a real codebase. 💡 Demonstrates: experimentation, architectural iteration |
🛰️ TraceRouteX
A custom traceroute implementation built from scratch — raw sockets, ICMP packet crafting, and TTL-based path discovery. A deep dive into how the internet actually works one hop at a time. 💡 Demonstrates: low-level networking, socket programming, systems thinking |
A full compiler implementation — lexer, parser, semantic analyzer, and code generator. University-level compilers project exercising the complete language-processing pipeline. 💡 Demonstrates: language theory, AST design, compilation pipelines |
|
Academic simulation project from K. N. Toosi University — modeling system behavior through event-driven simulation techniques. Bridges CS theory with real-world modeling. 💡 Demonstrates: simulation design, modeling, academic rigor |
🤖 MLTrain
A personal machine learning training ground — notebooks, experiments, and implementations covering supervised/unsupervised learning, model training, and evaluation. My path into AI engineering. 💡 Demonstrates: ML fundamentals, experimentation, data-driven development |
💻 LeetCode
My personal LeetCode solutions archive — problems solved across data structures and algorithms, spanning arrays, trees, graphs, DP, and more. Continuous sharpening of problem-solving skills. 💡 Demonstrates: DSA mastery, pattern recognition, algorithmic thinking |
My Codeforces training vault — curated problem sets and solutions used to prepare for ICPC (National Rank 20 🏅). Structured practice across difficulty tiers and categories. 💡 Demonstrates: competitive programming, contest prep, systematic training |
|
Implementations and notes from the legendary Abdul Bari Algorithms course — covering divide & conquer, greedy, dynamic programming, graph algorithms, and more. A classic DSA deep-dive. 💡 Demonstrates: foundational CS knowledge, self-directed learning |
A hands-on TypeScript learning repository — covering types, generics, decorators, advanced patterns, and idiomatic usage. Personal reference for building TS fluency. 💡 Demonstrates: type system mastery, language fundamentals, teaching mindset |
| 🧠 AI & Machine Learning | 🏗️ Systems & Infra | 🎯 Mastery Path |
|---|---|---|
| 🦜 LLM Fine-tuning & RAG | ☸️ Kubernetes Operators | 🌐 Distributed Systems |
| 🗂️ Vector Databases | 🕸️ Service Mesh (Istio) | ⚡ Event-Driven Design |
| 🔄 MLOps Pipelines | 📊 Observability Stack | 🏛️ Domain-Driven Design |
|
DEGREE B.Sc. Computer Engineering |
INSTITUTION K. N. Toosi University of Technology |
YEARS 2022 — Present |
GPA — / 20 ✎ |
✎ Insert your exact GPA above (e.g. 17.50 / 20 or 3.65 / 4.0)
My research curiosity lives at the intersection of scalable backend systems and applied machine intelligence.
|
🧠 Applied Machine Learning
|
⚙️ Systems for AI
|
🏗️ Distributed Systems
|
Formatted in academic citation style. Replace placeholder entries with your actual work as it publishes.
[1] Shadkam, P. "Image Processing Pipelines for Early Cancer Detection: A C++ Algorithmic Approach." Technical Report, K. N. Toosi University of Technology, 2025. [ preprint · code ]
[2] Shadkam, P. "Hexagonal Architecture in Production NestJS Services: A Case Study on Migrating Legacy Systems." Working Paper, 2024. [ code ]
[3] Shadkam, P. "Event-Driven Microservices for High-Concurrency Workloads." In preparation. [ expected 2026 ]
[4] Shadkam, P. "Competitive Programming as Algorithmic Training for Software Engineers." Student Essay, ICPC Training Notes, 2024. [ notes ]
— Proposed Thesis —
"Scalable AI Inference Architectures: Bridging Domain-Driven Backend Design with Production-Grade LLM Serving"
Advisor: [ To be confirmed ] · Department: Computer Engineering, KNTU · Expected: 2026
Abstract: This thesis investigates the architectural patterns required to serve modern AI workloads within clean, maintainable backend systems. We explore how Domain-Driven Design and Hexagonal Architecture — proven in traditional enterprise backends — can be adapted to the stochastic, latency-sensitive nature of LLM inference and retrieval-augmented generation. The work combines systems engineering with empirical evaluation of serving strategies under realistic load profiles.
Keywords: LLM serving · microservices · hexagonal architecture · MLOps · distributed inference
Relevant subjects from my B.Sc. Computer Engineering curriculum at KNTU.
| 🧮 Core Computer Science | 🖧 Systems & Networks | 🤖 AI / Data |
|---|---|---|
| · Data Structures | · Operating Systems | · Artificial Intelligence |
| · Algorithm Design & Analysis | · Computer Networks | · Machine Learning |
| · Discrete Mathematics | · Compiler Design | · Probability & Statistics |
| · Linear Algebra | · Database Systems | · Numerical Methods |
| · Advanced Programming (C++) | · Computer Architecture | · Signal Processing |
| · Software Engineering | · Simulation & Modeling | · Image Processing |
📐 ICPC Training Lead · KNTU Competitive Programming Team · 2023 — Present
Mentored junior teammates on algorithm design, problem-solving strategies, and contest preparation. Curated training problem sets leading to the team's ICPC National Rank 20 finish in 2024.💻 Backend Engineering Mentor · Peer Study Circle, KNTU · 2023 — Present
Led informal workshops on NestJS, clean architecture, and Domain-Driven Design for undergraduate peers transitioning into backend roles.📘 Open-Source Educational Content · Self-initiated · 2023 — Present
Maintain the public TypeScript Tutorial and Abdul Bari Algorithms repositories as learning references for the student community.
| 🏆 | ICPC (International Collegiate Programming Contest) — National Finalist, Rank 20 | 2024 |
| 🥇 | National Computer Olympiad of Iran — Olympiad Diploma | 2021 |
| 🚀 | Divar "T-Shape" Engineering Program — intensive training in high-scale systems | 2024 |
| 🧑💻 | Codeforces Online Contests — regular participant; see training archive | ongoing |
| 📜 | Additional workshops — fill in as applicable (research seminars, AI summits, hackathons) | ✎ |
Available upon request.
For academic references, research advisors, or letters of recommendation, please reach out via parsashadkam2004@gmail.com.
✎ If you'd like specific professor names listed publicly, paste them and I'll format them here.
🏆 I debug with the same intensity I solved ICPC problems — first principles, always
⚡ Started with C++ for competitive programming, fell in love with systems at scale
🤖 I believe the best AI engineers are great backend engineers first
☕ Coffee-fueled, terminal-obsessed, and architecturally opinionated
🌱 Always learning — currently diving deep into LLM orchestration & MLOps
🎯 I've shipped systems that serve thousands of concurrent users — and I'm just getting started
💙 Made with passion by Parsa Shadkam 💙