"LALAL.AI is hands down best for audio extraction quality, minimal artifacts & ease of use"

From DJ decks to radio studios and university classrooms, Francesco Cadente has built a career that blends creativity, technology, and teaching. Today he directs live radio at RDS in Milan, teaches the next generation of music managers, and still finds time to experiment with mashups and remixes.

We spoke with him about his early start in music, what makes radio such a unique medium, and how tools like LALAL.AI fit into his daily workflow.

“For me, radio is a world of its own, especially when you’re live!”

Francesco works across several fields, such as radio live direction, DJing, and teaching. We asked him how he got into the radio DJing and how these roles complement each other in his day-to-day life.

My journey started very early—I began DJing when I was 15. At first, like for many others, it was just a hobby. I always say that if I met my younger self, he would probably say, “Weren’t we supposed to become an architect?”

While I was studying to be a surveyor, music was just a real hobby but during the last years of school, I started liking it more and more. I began exploring genres and opportunities and got into the world of music production.

I’ll always be grateful to my parents, especially my mom, who encouraged me to check out the SAE Institute’s open day in Milan, right after I failed an architecture university entrance test.

So I stepped into a whole new world: from a small town like La Spezia, I moved to Milan and took courses in Music Production, eventually earning a Bachelor in Audio Production. My thesis project was of course, launching a radio startup. That’s when I subconsciously realized I wanted to be a DJ, but I wanted to do it on the radio.

Eventually, I made it to RDS Radio Dimensione Suono in Milan, where every day I combine music, technology, and creativity, always trying to learn something new.

Each of my roles enriches the others. Being a DJ keeps me up-to-date with musical trends; live radio work sharpens my technical and real-time decision-making skills; and teaching helps me reflect on my experiences and communicate them clearly, creating a virtuous cycle of continuous improvement.

Here’s a practical example: sometimes I have to replicate my radio directing work in front of a live audience, which is a totally different experience from being in a studio with just the hosts. But having a DJ background, literally from the ground up, gives me a real edge in any situation.

As for teaching, I get a lot of inspiration from the hosts I work with every day, especially in terms of communication and presence. What I teach often comes from my own experiences, which also prepares students for the various real-world situations they might face.

I teach in the Master of Music Management at Master Generation, and my goal is to give students the foundations of the music industry, but also help them understand how to behave and what values to have as future music managers, especially when it comes to dealing with mass media like radio.

And here’s the best part: in an era where “you have to know how to do everything,” I get to share insights they’d never expect and sometimes they give me fresh ideas too, both about music and tech!

I dedicate part of my lessons to AI; I think not talking about it today would be ridiculous and counterproductive.

I mention various tools, but I have to admit, when I show them LALAL.AI, their jaws drop haha!

How a typical radio DJ workday looks like

In the past few years, I’ve had the chance to work, and I’d say collaborate, with shows like “Peggio + Peggio” and “Guerrini Lanfranchi Show.” Some weeks, I focus more on one show than the other. In both cases, I always prepare the rundown: I analyze the “talk” segments, where we’ll run features and discuss topics with the hosts, and I check everything editorially, like songs, ads, promos, and news.

Then I prepare the materials for the broadcast or we share tasks among colleagues to make sure nothing is left incomplete. I check the equipment, and we’re ready to go live.

Once we’re on air, we follow the schedule but improvisation is always around the corner. You’ve got to be ready, like when an important call comes in or, recently, breaking news about the Pope.

After the show, I handle the editing for the episode and upload it as a podcast. Then I hand over the shift to my colleague.

On the radio, I mainly work on editing program podcasts. For editorial reasons, we cut only the vocal parts, removing ads and music.

I also often create background music for the hosts’ talk segments to help frame the discussion. Given the number of features we run, I frequently produce theme jingles.

For the past three years or so, I’ve also been making mashups for the “Guerrini Lanfranchi Show” every Friday.

Outside of radio, I’ve worked on other podcasts, handling sound design, mixing, and mastering. Sometimes I make mashups just for fun, and if they’re good enough, I submit them for broadcast.

I also spent time creating background tracks for YouTube creators, especially considering copyright restrictions and more recently, I’ve gone back to producing music for myself, especially bootleg remixes, with the goal of building a personal sound for my live shows.

Radio is a unique medium: nothing compares to it, not even streaming or podcasts.

Today, radio is also visual, like RDS’s Social TV, but audio is still the star. And it’s audio that sparks the imagination and makes people wonder: “What must it be like to work in radio?”

"Being into trends, it was only a matter of time before I saw a creator talking about LALAL.AI"

I came across LALAL.AI a couple of years ago, maybe even more! Being into trends and radio audio on TikTok, it was only a matter of time before I saw a creator talking about it.

And what made me want to try it? Curiosity! I couldn’t believe a service like that existed!

I told myself: I HAVE TO TRY LALAL.AI RIGHT NOW!

My first project [I used LALAL.AI for] was mashups, which I then brought to radio as a feature. And guess what? Some of those mashups were uploaded to the radio’s social media and went viral internationally! Crazy, right?

@rds_radio 🙌 Mashup da ascoltare e riascoltare 🔊 Un po’ di #Sanremo2023 e un po’ di storia: @Lazza x Gala x @sonolamadame per iniziare il weekend alla grandissima 🔥 @Claudio Guerrini @Roberta Lanfranchi #GuerrinieLanfranchiShow #mashup #mashupsongs #Cenere #IlBeneNelmale #FreedFromDesire ♬ Lazza x Gala x Madame Mashup RDS - RDS

Then I started creating short versions of tracks for both live and radio, super useful for segments and more. And lastly, still in the mashup world, I’ve built folders full of acapellas that I use during live sets to get the crowd singing. It’s a blast!

Sometimes I’ll play a super pop song’s acapella—it sounds like they’re singing with the original—and then boom, I drop the original instrumental or a remix for a big moment!

Audio cleaning (Voice and Noise) is also a big deal. Sure, there are systems and plug-ins that do this, but sometimes they’re pretty expensive.

Here’s a real-life example from the radio: Listeners often send us voice messages on WhatsApp, but not everyone has the latest phone or skips the car speakerphone. So the audio quality isn’t always the best, and sometimes we can’t even play them on air. But when the message is good and it’s just a bit noisy, LALAL.AI saves me!

I can clean it up perfectly, edit it like a pro, and make sure listeners, whether in their cars or elsewhere, understand it clearly.

Beyond that, it really makes a difference when I need to extract vocals from older songs that don’t have a remastered version.

"LALAL.AI is hands down best for audio extraction quality, minimal artifacts & ease of use. It’s become my go-to tool"

LALAL.AI saves me tons of time, the audio quality has gotten way better, and it lets me experiment more with music.

There’s one thing I find really important: the app itself.

Having accurate naming and a complete database of processed tracks is essential. I work a lot with Ableton Live, and having my LALAL.AI folder ready with all the separated stems, properly named, is a real win.

I don’t want to sound repetitive, but mashups are the most commonly used thing, mainly because they give listeners that WOW effect.

For example, sometimes when a song ends, I’ll take the instrumental version I previously separated with LALAL.AI and let it play out as the hosts begin their next topic. It really sets a great mood.

Besides, I used LALAL.AI to prepare remixes for live performances and club gigs. For the past two years, I’ve been adding two or three new tracks every week specifically for that purpose. Honestly, not having them nowadays would be a big mistake.

"The moment I show students LALAL.AI and how I use it in real DJ sets, podcast editing, or even classroom examples, they're usually blown away."

During his classes, Francesco is open about using AI tools in music production and showcases his students some of the gear with real-life applications. We asked him if the students are interested or mostly skeptical about AI in music.

My students are generally very open to AI-based tools, especially when they see real, practical applications. Once they understand how AI can enhance creativity instead of replacing it, the skepticism usually fades quickly.

That said, there’s often some initial hesitation, especially from those who think AI might “automate away” human skills. But the moment I show them tools like LALAL.AI and how I use them in real DJ sets, podcast editing, or even classroom examples, they're usually blown away. It becomes clear that AI is just another instrument in their creative toolkit.

In fact, I’ve noticed that many students become even more curious and start asking how they can integrate AI into their own workflows, whether in music production, marketing, or content creation.

It always depends on ethics, though. That’s literally the key factor. In some areas, it’s undeniably a shortcut; in others, it’s clearly an opportunity. I’m thinking of generative AI that can create entire songs—that’s probably the most important case.

That’s why I keep saying ethics matters. If a small business owner uses an AI-generated song to promote themselves a bit on social media, I think that’s totally fair. They probably don’t have the budget to hire a producer, a musician, etc. But if a multinational corporation does it... well, that’s where the ethics start to fall apart a little.

"Many creators still don’t take advantage of tools like LALAL.AI. I notice it from the quality of their videos."

They really focus on the visuals (which makes sense), but audio still takes a back seat for many. Something that would be a game-changer for a lot of them is removing heavy echo and reverb. I have no idea why they don’t use tools like these.

Maybe they think that the noise reduction on their microphones is enough.

If you're thinking about improving their audio quality but don’t know where to start, start with spending wisely or getting good advice on what gear to buy. Consider the idea of working with a sound engineer or a producer to level up your content quality.

Or just call me and I’ll record it for you haha!

Follow LALAL.AI on Instagram, Facebook, Twitter, TikTok, Reddit, LinkedIn, and YouTube to keep up with all our updates and special offers.