About

I'm a Computer Engineering student at the University of Manitoba, and the founder of Lectura AI, an AI-powered study platform, and Prepl, a real-time AI interview preparation platform.

I try to follow the same loop for everything I build: identify a real problem, understand the people who have it, build a solution, put it in front of users, and iterate on what actually happens — rather than treating engineering as a purely academic exercise.

How I Got Here

  • I'm originally from Addis Ababa, Ethiopia, and moved to Winnipeg, Manitoba to study.
  • Before university, I volunteered as IT support for the Melu'e Foundation in Addis Ababa (2023–2024), providing hands-on technical help to 20+ staff and volunteers and documenting processes — my first experience solving real technical problems for real people.
  • I started at International College of Manitoba's University Transfer program in Engineering Science in September 2024, then transferred into the University of Manitoba's Faculty of Engineering in April 2026, where I'm pursuing a Bachelor of Engineering in Computer Engineering.
  • I currently work as a peer tutor at ICM, helping other students learn — something that directly shaped how I think about Lectura AI.

What I'm Building

Lectura AI — Founder & Full Stack Developer, June 2025–present

Lectura AI turns lecture recordings, transcripts, YouTube videos, notes, and PDFs into structured, interactive study material: summaries, notes, flashcards, quizzes, study guides, and an AI tutor. The idea came from watching how much time students spend organizing material before they even start learning it — I wanted one system instead of five disconnected tools. I've worked across the entire stack: product, frontend, backend, AI integration, auth, payments, deployment, and iterating on real usage.

Prepl — Founder, August 2026–present

Prepl is a real-time AI interview preparation platform, originally built at a hackathon and then rebuilt from the ground up. It uses Gemini Live for real-time voice and video interviews with adaptive questions and immediate feedback, built around LiveKit and a small set of specialized agents (interviewer, grader, screener) rather than a single call-and-response chatbot. I architected and built the initial product myself before being joined by co-founder Kobiljon, and we're developing go-to-market through the Build with Gemini XPRIZE program.

Beyond the Code

I've spent time in startup and AI programs that push past "can I build this" into "should this be built, and for whom" — including the Kiro Hackathon, Google Cloud AI Accelerate, Gemini hackathons, North Forge's PATHFINDER program, RampUp, and Idea Start/I-VIC, plus mentorship through the University of Manitoba and Stu Clark entrepreneurship ecosystem. PATHFINDER in particular pushed me to actually interview potential customers and test assumptions instead of just building.

Where I'm Headed

I'm continuing to deepen my Computer Engineering foundation while building increasingly ambitious AI-native systems — real-time AI interfaces, agent architectures, and eventually AI for languages that mainstream models underserve, including Amharic, Oromo, and Tigrinya. That's a direction I'm exploring, not a shipped product yet.

Elsewhere