Online courses have become a global gateway into the tech industry. Among them, few have achieved the reach of CS50 at Harvard University — one of the most popular introductory computer science courses worldwide, now taken by more than 7 million learners and amassing over 162 million views on YouTube.

Behind the course’s cinematic look and distinctive visual identity stands Maxim Shcherbakov — known professionally as Max Scherr — a CS50 video producer with expertise in cinematography, video editing and creative direction. 

His creative work, however, extends far beyond academia. Max produced the viral music video “Volna” for the band Daite Tank (!), which has surpassed 1,8 million views. He also served as a director, writer, and editor of the short films Summerville and Hodge-Podge, both recognized at international festivals including the Bushwick Film Festival, Exit 6 Film Festival and Waco Independent Film Festival, among others.

In this interview, Max Scherr shares insights into CS50’s production process, and discusses the fusion of cinema and education, as well as the impact of new technologies on video production.

Let’s start with your background. When did you realize you wanted to focus specifically on video creation?

My journey into video production began when I was about eight or nine. We had a family video camera, and I would take it out with my friends, grab some prop guns, and shoot short films inspired by first-person-shooter games. It wasn’t intentional back then — just kids playing around. I think the moment it crystallized for me was at fourteen, when I saw Christopher Nolan’s Interstellar on a huge IMAX screen. Its synthesis of story, image, and sound hit me so deeply that I walked out knowing this was what I wanted to do.

You’re now a full-time video producer for CS50, one of the most popular online courses in the world. In your view, why has it become such a phenomenon? 

CS50 became popular because we work incredibly hard to make the course as fun and engaging as possible. Too many classes simply present material — no matter how dry — and rely on the student’s motivation. That might have worked decades ago, but today young people are overstimulated by social media and entertainment. You have to earn their attention.

This philosophy is woven into Professor David Malan’s teaching style: even the hardest concepts are broken down visually, clearly, and with real-world analogies. Combine that with the production quality I, together with my team, deliver — multiple RED cameras with cinema lenses, meticulously chosen angles, dozens of lights, perfectly mixed audio — and each lecture feels like a movie. 

When you eliminate every technical distraction, viewers online feel as if they’re right there at Sanders Theatre with the students. That level of immersion is hard to ignore.

Could you walk us through CS50’s production cycle? Which stages are you personally responsible for?

Prep for the fall course starts weeks before the semester — planning a visual theme for the year and creating an intro sequence, the equivalent of an opening credits scene you’d see before each episode of Stranger Things or Game of Thrones. A couple of weeks before classes start, I go into Sanders Theatre to run connections, place stage elements, build and configure all cameras, and set up the control station. Everything is tested rigorously — once the semester starts, there’s no margin for error.

Lecture days are tough 12-hour shifts: setup, testing, the lecture, breakdown, ingesting media. I usually live-direct three to four lectures per semester. Then I take all the media home and spend a couple of weeks editing each lecture to perfection, passing it through QA, and delivering various versions for multiple platforms.

You’ve also worked on CS50’s releases at Oxford and Yale. Did your approach differ from how you work at Harvard?

CS50 at Yale had its own end-of-semester Fair where students showcased their final projects. Our team traveled to New Haven with equipment to film interviews and highlights, and I edited those for year-end releases on YouTube and social media. 

At Oxford, I created a promotional video to launch CS50 there — one of the most fun assignments I’ve had. CS50 uses Sesame Street-style muppets as unofficial mascots, including one modeled after Professor Malan. I shot scenes of the muppet exploring Harvard’s campus, then directed my colleague in the U.K. to film matching scenes at Oxford. I edited it all into a playful montage of the muppet “traveling” from Harvard to Oxford, set to Queen’s “Don’t Stop Me Now.” That project was a blast.

One of the major initiatives you were involved in was the CS50 Educator Workshop in Indonesia. Could you elaborate on your experience? What set this project apart?

That was a particularly memorable chapter. In 2023, Indonesia’s Ministry of Education partnered with Harvard to run a months-long CS50 Educator Workshop. Thousands of teachers across the country took CS50, and top performers were invited to a week-long training on how to “adopt or adapt” the curriculum for their own classrooms.

The event took place at Jakarta Intercultural School. I covered videography, photography, and media support. The workshop was a major success — even Nadiem Makarim, Indonesia’s Minister of Education at the time, attended some events. Over a hundred teachers were placed at the forefront of modernizing computer science education. We recorded hours of footage, took thousands of photos, and the enthusiasm from local educators was incredible.

You significantly expanded CS50’s reach and audience on social media. Could you walk us through your strategy, and which approaches tend to work best in EdTech today?

In 2022, I was given full responsibility for CS50’s social media. We already had a strong YouTube audience, but other platforms were largely neglected.

Knowing people increasingly cluster around their preferred platforms, I aimed to build a unified multi-platform presence. I expanded to TikTok, Threads, BlueSky, and Telegram. Then I made sure everything — visuals, tone, messaging — felt cohesive.

I launched new content initiatives based on what performs best today — most notably short vertical videos that playfully tie into CS50’s themes while highlighting that the course is free and reputable. Algorithms loved them. Over a few years, I created more than 150 Reels that collectively reached nearly 20 million views and brought in hundreds of thousands of followers. 

Overall, I think staying flexible and attuned to trends is crucial. The internet is a living organism; audiences want to feel that your content is part of it.

Turning to your personal work, Hodge-Podge stands out — a short film about a delivery driver who comes across a man attempting to end his life. How did this idea originate, and how did the production unfold behind the scenes?

By the time I came up with Hodge-Podge, I had been living in the U.S. for a while and had met all kinds of people from the immigrant community. My close friend worked in DoorDasher — a job many foreigners take early on. At the same time, I saw older immigrants who had lived in the U.S. for decades yet never fully adapted, sometimes barely speaking English. If you walk through a major city’s Chinatown or New York’s Brighton Beach, you’ll see what I mean.

These two extremes fascinated me — the very beginning of an immigrant journey and the very end — so I decided to collide them in a high-stakes, fast-paced situation.

I also knew I wanted to make the film “properly,” with a real production company and strong actors, which required money. I took out a loan, and the film ended up costing close to $40,000. We shot for three nights, and I spent a couple of months editing.

In your opinion, why did the film resonate so strongly with festival audiences? What makes it special? 

What I love about Hodge-Podge is that neither character speaks proper English. They can’t communicate verbally, but they don’t need to — empathy doesn’t require language. 

Festivals in the U.S. and U.K. responded strongly to the nonstop pacing that transitions into something tender by the end. I also intentionally left out subtitles — if the characters can’t understand each other, neither should the audience.

How do you typically assemble a team for your projects? Do you have regular collaborators?

Every project so far has had a different team — sometimes due to geography, sometimes due to circumstance. My only consistent collaborator is my wife Ksenia Novikova, known professionally as Senia Novak, a VFX artist and animator. We met in 2020 and have worked together ever since, including on CS50.

Visual effects are everywhere now; we truly live in a digital world. On my films and music videos, we’ve used VFX to enhance scenes that would be too costly or impossible to create practically. At CS50, Senia worked with me on virtual classroom environments using Unreal Engine and other tools. She knows my style and creative vision and can implement it flawlessly. We are quite a power duo that way.

In your view, how can the film industry and the education sector learn from one another to engage audiences more effectively?

Educational content should move closer to a cinematic approach, as CS50 does. It removes a barrier between viewer and material. But even if you can’t shoot with cinema cameras or complex lighting, you can still think more carefully about viewer retention — simplifying ideas, adapting to modern viewing habits, and adding moments of humor. A little comedy goes a long way.

How is technological development — including AI — influencing video production today? What opportunities and risks do you see? 

People often say that new, affordable video technology lowers the barrier to entry — and I agree, even if it also means more competition. One thing is undeniable: AI’s influence is massive. Even at this early stage, you can use it to enhance phone footage, clean up audio, fix small errors, or create visuals that would otherwise be too expensive to produce.

Video generation tools can bring wild imagery into your projects without an enormous budget. At the same time, audiences can react negatively when AI is overused, and the industry is still trying to understand where the boundaries should be. I tend to believe that if your intentions are good and your heart is in the right place, viewers will recognize and appreciate that.

I used AI generation in Hodge-Podge for shots I simply couldn’t afford to create otherwise. But new iterations of tools like Sora or Veo genuinely terrify me. All to say, I have no idea where this is heading. Your guess is as good as mine.