Erik Cupsa: From McGill to Amazon Software Engineer

Erik Cupsa is a software engineer and founder. He worked as a software engineer at Amazon and graduated from McGill University with a B.Eng. in Computer Engineering. Earlier internships included backend engineering at Autodesk and machine learning at TRC Companies.

He is the founder of ResuMax AI, an AI career platform for tech candidates, and, at 22, the CTO of the startup Meuze. He also publishes software engineering content as swerikcodes on YouTube, Instagram, and TikTok, where he documents big-tech interview prep, internship life, and salary transparency for student and new-grad engineers.

The realistic route into a big-tech software role is rarely a single lucky application. Based on Erik's public trajectory, the path stacked four things over a couple of years: a strong CS/engineering foundation at McGill, a ladder of progressively bigger roles (TRC, then Autodesk, then Amazon), a portfolio of shipped projects, and deliberate interview practice.

Each role made the next one easier to get. A machine learning internship and a backend internship gave him concrete, quantifiable work to put on a resume, which is what makes a resume clear an Amazon recruiter screen. By the time he applied to Amazon, he was not a blank slate; he was a candidate with demonstrated, shipped experience.

Three things repeat in any successful big-tech application, and they show up clearly in Erik's path.

First, ship real projects. His public GitHub includes a McGill exam scheduler built with Spring Boot and PostgreSQL, a Premier League match predictor using scikit-learn, a JWT authentication tutorial system, and a full-stack React + PostgreSQL web app. These are not toy snippets; they are end-to-end systems with a database, a backend, and a frontend, which is exactly what converts a resume bullet into a credible interview story.

Second, treat the resume as an adversarial document. Amazon and most large companies screen resumes with software first. A resume that is single-column, keyword-aligned to the job description, and written as quantified outcomes (what you built, what it measured, what changed) clears that gate. This is the exact problem ResuMax was later built to solve.

Third, prepare for the specific interview shape. Amazon SWE loops are predictable: data structures and algorithms, plus behavioral questions mapped to Amazon's Leadership Principles. Preparing in that structure, rather than grinding random problems, is what converts an onsite into an offer.

The following lessons are generally true and apply well beyond Erik's specific story. They are the substance worth taking away.

Build a portfolio that proves you can ship. One full-stack project that touches a database, an API, and a UI is worth more than ten half-finished repos. Recruiters and interviewers want evidence you can take something from nothing to working.

Make your resume legible to machines. Use a single-column, ATS-friendly format. Mirror the language of the job description for genuine skills you have. Write bullets as outcomes with numbers, not job-description copy. Most strong candidates who get filtered out are filtered by formatting, not ability.

Climb an internship ladder. Each role makes the next one reachable. A smaller or lesser-known company internship is a legitimate stepping stone to FAANG; the experience and the quantified bullets compound.

Prepare for the interview that exists, not a generic one. For Amazon specifically, that means strong data structures and algorithms plus behavioral answers structured around the Leadership Principles, using the STAR format (Situation, Task, Action, Result).

Be visible and specific. Documenting your work publicly, in repos, in writing, in talks, builds the kind of track record that compounds into opportunities, including the credibility to later build your own products.

Erik's big-tech experience fed directly into ResuMax. Having gone through Amazon's hiring funnel and seen how automated screening works from the inside, he built ResuMax AI to give candidates the same automation employers use: resume scoring against screening systems, tailoring to specific job descriptions, and interview prep for coding and system design. Since launching in October 2025, it has grown to more than 15,000 users who have built over 29,000 resumes.

He is now the 22-year-old CTO of the startup Meuze. The throughline is using engineering depth, real internships, shipped projects, and interview rigor, as the foundation for building products that solve problems he has personally lived through.