Big Cloud Connect: An Interview With: Hazel Savage; CEO of Musiio
In the first of our new Big Cloud Connect ‘An Interview With…’ series, our recruiter Jack sat down with CEO and Co-Founder of Musiio, Hazel Savage. Based in Singapore, this tech company is the first venture capital-backed AI music start-up in Southeast Asia.
In this get to know you with Hazel, we find out what it’s like to be a woman in tech, how she hires top talent and what’s next for Musiio following their successful round of seed funding.
Jack: So Hazel, you’re from Musiio. What is it and what does it do?
Hazel: I am indeed. Hi. Hi Jack. Thanks for having me. So my name’s Hazel Savage, I’m the CEO and co-founder of Musiio, an artificial intelligence company for the music industry. We’ve been a company for about a year and a half to date. And essentially our company is a combination of two very specific skills that myself and my co-founder have. So I’ve been in the music industry for about 14 years. Previously at Shazam, Pandora, Universal. And my co-founder has a background in data science, building artificial intelligence and computing. So really we sort of came together from the idea of ‘can I identify a problem that the industry has that’s hard to know if you’re not an insider’ and ‘can he build something that this industry doesn’t have the capability of building for itself.’ So we kind of started with that premise but then really came up with the idea that, you know, when I was – actually I love talking about it. We’re talking over 10 years ago, so when I had my very first job I worked for HMV in the UK. And I, you know that was my Saturday job while I was at uni and I used to just put the new release singles out on the shelf every, every Sunday night ready to buy on Monday. And on a really quiet week there’d be maybe 2, and then on a really busy week maybe like 5 CD singles. But then you know – So that’s a very manageable amount of music. But you switch forward to today and even Spotify announced that they now have 40,000 new tracks a day uploaded onto their platform. So – and I mean that’s just one of the platforms. You know, you look at UGC content (user-generated content) and the platforms like you know Bytedance, Tik Tok, SoundCloud – all of the others – and you look at huge volume in growth. So essentially what you’re looking at now is you’re going from a very small volume of music that was very easy to manage, right through to a very high volume of music that’s now physically impossible to manage manually. No-one can listen to 40,000 songs a day. In fact, you could probably hire a thousand people and you still couldn’t manually listen to everything. So that’s really the challenge that we operate on and we do that by an automated AI tagging, automated AI search and automated AI playlisting.
Jack: So I suppose even if you did get a thousand people to listen to 40 songs each there’d be a lot of opinion from each individual on what music is about or what’s in it. Whereas an AI will make quite a linear decision about things based on what the rules are that you’ve put into it.
Hazel: Yeah and that’s true as well, I mean even if it was physically possible I think the average human maxes out at about 200 tracks a day manually. I know because I’ve had people work for me and I’ve done that myself. I’ve manually tagged tracks. But the real challenge is not just that, but the human is affected by things all across the way. So whether I’ve had my morning coffee maybe I’ll be less generous about the stuff I’m listening to. Have I had my lunch yet? Am I quickly trying to get to a meeting? Do I need the bathroom? You know, all these other factors that affect. You know, I might be having a bad day so I’m giving everything a very low score and I’m not very happy. Or I might be really happy with everything. I might have listened to 5 bad tracks in a row, so the 6th track I give an amazing, amazing review or an amazing rating, really generous tags because the last 5 were so bad. An AI doesn’t have any of that bias, so that’s a real benefit to it too.
Jack: Okay. So, you guys picked Singapore, I’m guessing you’re not from Singapore?
Hazel: I’m not. You can tell by my accent, can’t you? Yeah, so I’m from the UK originally. A place called Darlington, so up near Newcastle. I actually haven’t lived in the UK for about 9 years. I was in Australia for 5 years and I’ve been in Singapore for 4 years now. And so I was here for work originally, working for some other start-up and a year and a half ago decided to try my hand at starting my own business. So I was kind of in Singapore already, but it turns out that Singapore is a great place to start a business and a really good place to build a company.
Jack: Yeah, especially a tech business.
Hazel: Yeah, I mean the talent here as well – you know they’re probably all listening to me now as I’m sat in the office as I’m talking to you! But yeah, we’ve found that actually as a company, Musiio was the first-ever VC funded music tech company in Singapore. If you look up where other funded music tech companies are there’s obviously a hub in London, there’s quite a lot in LA. Austin and Nashville are becoming quite big. But really in Singapore, we’re kind of the only one that have raised money and decided to build in Singapore. So there’s disadvantages and advantages to being the only company. I think one of the advantages has been our access to talent. You know there’s a lot of people – we’re not competing with other companies to share a voice. You know it’s not like they can – There’s not a development team here for Spotify. There’s not a development team here for you know 2 or 3 other exciting music start-ups. So we have that kind of individuality that separates us out. But the real challenge or the company we do compete against would be like the banks or the FinTech companies. There’s a lot of finance over here. Which again is something that you know depends – completely dependant on the candidate. But we’ve been really fortunate to find people who wanted to work for a company they were really passionate about, wanted to work for a company they were really interested in and believed in the mission and that’s enabled us to hire really well in this region.
Jack: You know that’s always important for start-ups, isn’t it? Finding the people that really buy in and give that extra commitment I suppose when you’ve not got 4000 people to do work that one person that does this, one person that does that. They’re so much more important – Find those that really believe. You’re right as well with – You’re actually one of the first people that’s said it that way round. That we’re competing with banks. A lot of start-ups say you know, we’re not competing with banks because we’ve got a totally different product. But you both want people with Python, you both want people that have used Go or Google Platform. Whatever it is, so you are competing with each other regardless of your product. Because you’re all speaking the same programming language.
Hazel: Yeah and I agree. I think you need to be a bit of a realist and I think you need to be to say well, you know people work to earn money. So one of those factors is always going to be salary. And one of the biggest challenges in a start-up is that you very rarely have the ability to pay at the same level as like a banking institution. So then you need to think about well, you know obviously ideally we’re looking for people who we can pay fairly but are motivated by more than just a paycheck and then also perhaps more interested in the product. You know there probably is no end of backend banking and login systems and money movement systems for any of the huge players in Singapore. But the chance to work for Singapore’s first music start-up I think is probably a little bit attractive especially for people who are particularly passionate about music. Perhaps they play instruments as well and they want to do something that they love. And then the other thing that I think we can do to help compete is we like to bring the best of kind of English and then kind of Swedish workplace practices to our company. So we have a more generous holiday allowance than a lot of the traditional companies here, we’re very flexible in terms of working hours, it’s not a corporate dress policy. It’s a very, very flat structure for our hierarchy and various other things that you know I think I’ve sort of taken the best of the things that I like that the companies I’ve worked at over the years such as at Shazam and Pandora. And then really brought the best things into Musiio to kind of help build what kind of a company we’re going to be.
Jack: Good. My opinion is that people should always focus on the company culture regardless of the size of the business. I don’t think there’s many truly mercenary candidates out there that just want to go wherever pays the most. I don’t meet many of those people. It’s just the circumstance that people are in. Some people are able to give some time like this, others have other commitments in their lives and they’re just not at the same stage. So it’s about the two coming together usually. But companies that focus on the culture do much better.
Hazel: I completely agree. I think there’s a huge focus on you know why someone would want to work here and how they would enjoy it. That’s hugely important for companies of any size. And then also thinking about that kind of candidate fit. I personally have always worked at either start-ups or you know, new offshoots of existing businesses because I consider myself a kind of – I’m a very sort of 0 to 1 person. But then you know usually it’s someone very different who’s your 1 to 100. And so you know there are people that prefer the structure of corporate. There are people that prefer the kind of the rules and the process and having everything being really, really stated. And then there are people who prefer kind of building the path as they’re walking along it. And I think trying to find the right candidate for the right company fit is something else that’s really important. Important when, when you know, we’re interviewing and important as you scale the business because as well we’re kind of right at that size where I think I’ve probably interviewed everyone that’s here now. But you very quickly get to a size where that’s not possible. So how do you ensure that carry over once you’re past a certain size as well?
Jack: Yeah definitely. Well, that leads me into asking about where the company is at right now in general with the products, how that’s performing, or whether that’s out there yet. Funding and what you’re looking for next. Where is it at now and where are you going in 2020?
Hazel: Yeah absolutely. So in January of this year 2019, we raised 1 million USD in funding from VCs. And so I would be looking to raise more funding, so probably about mid next year. We’re a team of 14 people all based in Singapore and we over this last year now have 12 enterprise customers. So that means we do a lot of large scale artificial intelligence, a lot of custom integration and a lot of custom delivery. A couple of customers that I can talk about who we spoke about in the press – we work with Audio Networks who are really great sync company. So sync the licensing of music for film and TV. They have a catalogue of 170,000 songs, we’ve given them an AI tool that enables search on their entire catalogue very quickly using an audio reference search. So the ability to drop a track in and then in 2 seconds have it search 170,000 songs and then pull back the 10, 20 most relevant tracks in a couple of seconds rather than having to kind of think well what have we got in the database, let me find this person and where’s that track and who did that last week and maybe try some keywords. So basically AI-powered tools for the Audio Network. And they have 9 offices globally so they’re a really great one that we’re working with. And also another great company called Overcoast who are a boutique agency – offices in London and the US. Who basically needed us to tag their catalogue. So again where you know even if a catalogue is 5, 10,000, 20,000 tracks, doing it manually is no longer the most cost-efficient or accurate way of doing it. So we work with them on their catalogue tagging. So yeah. And a lot of others that are super exciting that I can’t talk about quite yet but it’s all going well.
Jack: Alright, it’s not that type of interview I won’t push you don’t worry.
Hazel: It’s a hard journalistic interview with lots of press!
Jack: Yeah, I’ve just done Louis Theroux’s book so I should now be an expert at pushing people into giving me really secret answers.
Hazel: Got it. Okay, good.
Jack: In terms of something that I wanted to ask you about we touched on a little bit before with the competition in Singapore I suppose, but with you as a business leader and with spotlights with things like women in tech, what are your thoughts on opportunity especially in machine learning and in the areas of data engineering in Singapore or your experience anywhere?
Hazel: Yeah I mean it’s a, it’s a tough one as well because even though I’m the CEO of a tech company I’m not the technical co-founder. So you know I find that having a good understanding of the technology is absolutely essential. I need to be able to understand what we do so I can explain it and sell it to other people. But I can do that without being able to code myself. I think it’s really interesting actually because I organized a dinner the other night and I, you know I was like I’m just going to invite all my favourite CEOs. Because I met my co-founder through a start-up incubator called EF. And so, you know, EF definitely had more men than women. That’s just how it goes when it comes to applications. And then when I – you know – I found when I was putting together my shortlist I was like, oh there’s not as many women on here as there should be you know. I’m trying to sort a nice sort of 50/50 balance dinner party and I was really struggling to find female CEOs to invite apart from myself. And so I think it is something that I feel I want to be proactive about to make sure there are opportunities. I think it does as an industry, as an entrepreneur and a founder, it definitely does still skew towards there being more men. However there are these great kind of enclaves where women get together and exist so, there’s a really great online community called Elpha which was started by two women that used to work at Y Combinator in the US and I love that. It’s an online community. And there’s also one in Southeast Asia called She Loves Tech run by a woman called Leanne. And again, you know, these opportunities where we come together and we work on stuff together are really important. And I think I guess technology more than anything has made it really instantly accessible for me to meet Leanne through Instagram, for me to be a part of the Elpha community without being based in the US. I think the reality of the day to day it’s still – there still seems like there’s a lot more men that are founding starting companies. But I was really pleased as well because I was having a chat with our lead investor the other day and he said that 25% of all their founders that they’ve funded are women. So I mean that’s a really high percentage given the kind of global standard. So I think it’s changing, it’s out there, but you’ve got to do a little bit of discovery to kind of find the communities and find where they are. But I’m happy to go looking.
Jack: Were you finding in Singapore with – when you’re interviewing I’m guessing you get a lot more applications from men because there’s just more men at the moment in data science in Singapore? But are you thinking about how you could try and make it more equal in your own company? Have these things cropped up?
Hazel: Yeah, I mean I think, I think you’re absolutely right. You know, that the fact of the matter is if we put up a role for data science or a developers role, the chances are that we’re going to get something along 90% plus applicants are going to be male. And I think interestingly as a company though you know we’re sort of, relatively diverse and in terms of along gender lines but also the other elements to factor in, in Singapore is diversity along racial lines. And it’s really important to me to have you know not just sort of white European representation in the company but to have Chinese Singaporean, Indian Singaporean, Malay Singaporean – and also people from other countries as well. Russia, India, Sweden like my co-founder. And to really have diversity across the board and that’s not always just a male, male-female thing. In Singapore, it’s sometimes more complicated and more important to make sure you have a lot more diversity. But yeah, you know we’re still relatively small so all it takes is 1 intern out, 1 in for the balance to now be shifted to 50/50. But you know, I think, I think as a female founder I put more focus on hiring women and probably more focus on hiring a diverse team. And I think that’s fine. I think all you have to do is just look at every CV as it comes in on merit but also not discount your own prejudice and try and interview a variety of candidates. It’s working well for us so far you know, we have really good representation across the whole team.
Jack: Yeah. Good. I find a lot of people – managers in Singapore ask me if we can try and focus on representing more female candidates and trying to balance up teams a little bit more. We were talking then in relative terms because when you say 9 out of every 10 applicants are male it becomes unfair to ignore 9 out of 10 people that have applied for a role. Obviously, that’s never going to happen. But Singapore relative to the rest of the world is still really progressive. It’s one of the reasons that we love Singapore, it’s that both for gender and racial diversity Singapore are still an amazingly open country, amazingly innovative, because it’s got a lot of different backgrounds and cultures. With Singapore the best of that I think it’s going to get even better over time until you reach capacity. Singapore is still really small.
Hazel: I think you’re right there, what you said is you know, if you’re asked to have a focus because the team is trying to hire more women that doesn’t mean – yeah you’re right – doesn’t mean ignoring 9 out of 10 CVs that they’re receiving. It means, put a few great women forward, put a few great men forward. They’ll end up hiring the best candidate out of the top selection that you pick. And as I say I think that’s really the best way to ensure you have a diverse team. I think you know, again I’m not quite sure what sort of size company but I think smaller companies sometimes have even more of an issue because it can be quite common for start-ups to get to kind of 5 or 6 people or even up to like 8 or 9 people and still not have hired a single woman. Which I think is a little problematic. But then again, it’s hard to know obviously I don’t work at those companies so I don’t know who applied or what their processes were. But I think yes just to draw a little bit of attention. You know I think if and when I was working with a recruiter you know, I would not expect to see the top 5 CVs recommended to me make to be all men and to be all of the same race. I would expect some diversity. And I think that’s just got to be your starting point. Hire the best candidate but be open to meeting multiple candidates.
Jack: Yeah. Just to take your idea, by the way, it’s the same in recruiting. We have the same issues with – it’s a very male-dominated industry. So we’re not the answer to it all but it’s important to draw attention to things like this. Finally then Hazel, what’s next for Musiio? Are you guys aiming at anything in particular at the moment?
Hazel: Well, actually in the music industry the Christmas period is probably from about 1st of December to the end of Jan is a particularly quiet time because a lot of people take leave so I’ve got to – that’s not going to stop me though I’ve got a ton of stuff to be working on. You know all the admin tasks that I’ve been leaving till the end of the year which we thought would be quiet. But for us, it’ll be about closing as many deals we can before the end of the year. You know having a nice sort of Christmas celebration with the team in December. And then you know, keep working hard early next year until I have to start pulling together to do a Series A. So yeah, that’s kind of the plan for the next few months.
Jack: Alright well, I wish you luck. I think you’ve got a product that’s got a niche which is so important. There’s not a ton of people doing it. The solution is useful. I’d like songs to be recommended in a better way. It’s offensive when I listen to a subscription service and it recommends music that and I always think you know, music’s very personal to people isn’t it? When somebody recommends something that’s just completely outside of what you want its – it can feel as if somebody that you’re paying doesn’t know you very well. Even though I know it’s an algorithm. I’d like you to do well just because I’d like the music business to be able to do better and I think that it will help lesser-known artists get their music out there. Like you said there’s so much content we’re never going to listen to it all. And now none of us buy albums anymore we all just skip to 30 seconds through thinking does that song sound good, I’ll give 3 songs off this album a try and then not listen to the rest of it. So it would be nice if there was a solution to that because I miss listening to whole albums. I’m starting to feel old, I’m only 32.
Hazel: I mean I’m even older than that. But also, I read a really fascinating article the other day that your music taste is really just the music that you discovered when you were 14. And I would say that’s probably pretty accurate? I don’t know that I’ve massively updated what I’ve listened to since I was 14 which is a long time ago. So yeah, but the oldies are the goodies and I’m perfectly happy with that. There’s never a judgement on my part.
Jack: Yeah, I’m trying to listen to more music. There’s too much of it. Hopefully Musiio can help with that. Well, thanks, thanks again. Really appreciate it. You are our first person. So you’re very very special.
Hazel: Excited for it. Thank you for having me!
We feel so privileged to get up close and personal with some of the most inspiring figures in data science, machine learning and artificial intelligence – and Big Cloud Connect makes it possible for us to do so!
If you have a topic you would like to discuss with us or you want to share your experiences in the industry don’t hesitate to contact us here.