I’m Youssef Elsawy, a computer engineer who designs AI agents and automation that plan, act, and check their own work.
Arrow keys move between nodes.
Simulationruns in your browser, no AI model is called
Select an agent to run a task
Most software waits to be told what to do.
I build the kind that works out the next step, takes it, and checks the result.
That takes engineering, from the circuit to the cloud, until it runs on its own.
I’m a computer engineer drawn to the places where software meets the real world: autonomous AI agents, automated workflows, embedded systems and robots that have to behave reliably outside a demo.
My approach is to treat intelligence as a systems problem. Clear interfaces between agents, tools that fail loudly, review steps that catch mistakes, and humans in the loop where it matters.
Long term, I want to turn ambitious ideas into products people rely on, and to build the company that ships them.
Agents get roles, contracts and observability, so behavior can be traced and improved.
Automation is only useful when it checks its own output. Review loops are part of the design.
From firmware and sensors to APIs and interfaces, I like owning the whole path.
Working systems beat impressive demos. I build to put things in people’s hands.
Case studies, each with the architecture behind it. Hatched entries are open slots for projects still being written up.
A live hiring platform where job agents find roles, early-career first, check the fit and draft each application, and by default nothing is sent until the candidate approves it.
A research agent that runs on your own computer, reads LinkedIn, X and the open web in your own signed-in Chrome, and turns a goal written in plain words into a ranked report with links. The reading costs zero AI tokens.
For ASU's AYCE Digital pilot program, I tied the course, completion, user and progress data kept in Excel sheets and queries, and the program's Salesforce case handling, into one automated workflow, replacing the manual typing and updates that used to happen every day.
How I make videos and ads with AI: a written brief becomes a storyboard, generated shots, a motion design pass and a finished ad in every format. Interactive motion demos run on the page.
A deterministic, in-browser simulation of a planner, researcher, executor and reviewer working through real-world tasks.
A two-wheel robot on a Raspberry Pi Pico 2, written in C, that holds each wheel's speed with encoder feedback and PID, and drove a straight 6 second run with its wheels 2 pulses apart.
One sentence: what it does and for whom
One sentence: what it does and for whom
Simulation
A deterministic simulation that runs in your browser, and no AI model is called. It shows how I structure multi-agent systems: clear roles, explicit tool calls, a reviewer that can send work back, and a human checkpoint where it matters.
“CI is red on main. Find the cause and ship a fix.”
Reproduce, diagnose, patch and verify, with a reviewer gate before merge.
The planner drafts the steps once the run starts.
Arrow keys move between nodes.
Tools in this task: Tools: Sandbox, Memory, Search
Use the arrow keys, Page Up, Page Down, Home and End to move between events. Press Enter to show or hide the payload.
Every message, tool call and decision appears here, with its timestamp on the simulation clock.
Multi-agent orchestration, tool use, evaluation and review loops, and workflow automation that runs without babysitting.
Full-stack products from data model to interface, built to be maintained.
Computer architecture, operating systems and networks: how the machine actually runs the code.
Firmware, sensors and actuators, and the control loops that connect them.
The tooling that keeps systems shippable: version control, CI, containers and cloud.
Pre-university
Before university
I started by building Arduino robots in C and C++, then taught myself Python, SQL and Java.
Arizona State University
Studying Computer Systems Engineering at ASU, while building AI agents and automation of my own.
Work
W. P. Carey Executive Education, Arizona State University
Part of the team that manages the AYCE Digital platform and program. I automated its course, completion, user and progress data and its Salesforce case handling into one daily workflow.
Education
Arizona State University
Next, from Spring 2027: a master's in robotics and autonomous systems, with a focus on AI.
Ahead
The road ahead
Engineering work on AI agents, robots and autonomous systems, and in time a company of my own.
Contact
Open to engineering roles, collaborations and early-stage ideas, especially around AI agents, automation and intelligent hardware.
Prefer a form? Write here and I’ll answer by email.