Kiwi start-up Transcribe Me just closed an oversubscribed angel investment round, raising $1.2 million from private investors in New Zealand, the US and Europe (along with Ice Angels were investors from US-based syndicates including Tech Coast Angels, Sierra Angels and Maverick Angels). It's receiving rave reviews from Singapore to San Francisco for modernising and streamlining voice-to-text transcription. And its CEO believes monotonous hours spent typing out meeting minutes are numbered. Transcribe Me's Alexei Dunayev goes on the record.
Comparing the tech revolution of the past 20 years to the imagined future created by movie directors is both laughable and cringe-inducing. Where are the flying cars and teleportation devices we were promised? The era of innovation has provided many devices and services previously stuck in the realm of sci-fi movie fodder, and one such futuristic ideal is voice-to-text transcription. It continues to elude the tech industry and remain just that – an ideal. Until Transcribe Me came along.
The SaaS (software as a service) app is the brainchild of Kiwi entrepreneur and Fulbright scholar Alexei Dunayev and his international team. It seems so simple, in a Back to the Future view of technology. Using your smartphone, you record your meeting, interview or lecture, then hit the ‘transcribe’ button in the Transcribe Me app. Within 48 hours your recoding is emailed back to you as a Word document. Sure, it seems simple. but according to Dunayev, it's not simple at all.
“Despite having all the other technologies we envisioned we’d have 10 years ago, voice-to- text has something that eluded researchers and it’s quite a complex technical problem,”
he says. “Ever since I was a teen messing with computers, I was always looking forward to that future when you can talk into a computer and it knows exactly what you’re saying.”
Dunayev has an infectious enthusiasm for Transcribe Me, both as company and as product, despite it officially still being in the beta testing phase. His passion is apparent even through an unreliable Skype connection from a San Francisco shopping mall car park. Dunayev speaks in depth, with breadth, and at speed about his product and the high-tech industry for start-ups. As it turns out, it was exactly the kind of conversation, where upon finishing, I wished I could just hit ‘Transcribe Me’ and have returned from magical elves to me in a document form the next day.
Time to pounce
Dunayev knows just how far the technology has come. He felt it high time for Transcribe Me to pounce on the hole in the market, and despite the hype, he still believes Transcribe Me can evolve.
“The problem of being able to have a real-time, word-for-word, perfect voice-to-text has still not been solved. It’s one of the big technological challenges developers can’t perfect, so the model we put together is kind of a hybrid.”
The company’s model combines the best software speech recognition algorithms, with the accuracy of human transcribers.
Although the company’s intellectual property lies in the voice-recognition algorithms, Dunayev insists it’s his crowd-sourced transcribers that give it the edge over the competition. But to ensure quality and a quick turnaround, Transcribe Me wanted to cut out the middle man.
“We’ve very different to our competitors, we directly work with our transcribers – we don’t use a third party like they do.”
It’s that direct relationship with a global pool of transcribers that’s the key to Transcribe Me’s ace-in-the-hole: microtasking. The secret recipe for this innovation’s success is the combination of voice-recognition algorithms, global crowd-sourcing and then microtasking for those individuals who want to make a quick buck.
“Our service helps them monetise their downtime,” says Dunayev, as if he’s pitching to me as a potential employer. “Right now, if you look at crowd-sourced transcription, typically one of our competitors would go to their database and ask, ‘Hey, can anyone do this one-hour long conversation?’ and when someone says ‘Yep, I can’, they get given the whole job.”
The result is often a week-long wait for the customer while one transcriber works through the audio. Conversely, microtasking splits and then standardises one large laborious job into many small units. Combine microtasking with crowd-sourcing and as soon as you’ve finished recording that awkward guest speaker at your annual conference and hit the transcribe button, the hour-long audio file is uploaded to the company servers and chopped up into digestible 10-second files. Those files are then dispersed amongst the pool of transcribers, who take about one minute to distinguish every ‘Ummm’ from every ‘Ahhh’ in the recording.
The clips are then filed and rearranged, like a jigsaw puzzle, to recreate the recording into a Word document for the client. And this all happens within 48 hours.
Of course, 48 hours is just not good enough for Dunayev.
“We want to deliver transcripts within minutes, not hours,” the Stanford MBA graduate says. “We’re not there quite yet.”
He knows that it’s just a matter of constantly improving the algorithms and building on Transcribe Me’s existing assets before that becomes a reality.
“We can get hundreds, even thousands, of people getting little tasks in parallel,” he explains. “But you don’t need the whole audio to begin transcription – you can begin with the first chunk, give it to the computer to solve.
"If it can’t solve it, then pass it on to a human to solve.”
It’s the human element that gives Transcribe Me an edge over other voice-recognition apps (sorry, Siri). Its pool of talent has not only the ability to distinguish multiple speakers at meetings and interviews, but also the ear and experience to pick up on subtle inflections, exclamations and sarcasm frequent in human dialogue, adding to its precision.
Dunayev says you can have hundreds of people working in tandem on a single job, resulting in 98 percent accuracy.
But accuracy and efficiency aren’t enough for Dunayev. The price tag on the service was fair game, too. The app is free to download, but Transcribe Me undercuts its closest competitors by half, with prices starting at $1 per minute of raw audio.
The company bridges the global-local divide by matching local transcribers to the smartphone’s location, minimising the effects of local accents so transcribers don’t spend time scratching their heads when they hear ‘fush ‘n’ chups’ or add unnecessary punctuation due to our Kiwi habit of a raising intonation to make our sentences sound like questions.
Dunayev believes the sum is far greater than the value of each of its parts. He calls the ability to combine all these elements “the secret sauce,” but it’s not just clients he hopes will gain an appetite for the app. He’s very proud of offering ways to help people from around the globe monetise their spare time. With the lifestyle pressures of many people working more than 50 hours a week, it’s hard to believe that employees arrive home, sink into the couch and then jump at the chance to deconstruct a stranger’s rant. Dunayev is quick to explain the benefits of his company’s flexibility to his cynical interviewer.
“Our employees might be sitting at home watching TV, the ads come on and they can do five minutes of work. They could be on a bus, or waiting in line at a bank,” he says. “They can monetise the smallest chunks of
their downtime. No longer do they have to put aside five hours of their time to do a job, they can do it when it suits them.”
It’s therefore a flexible way to supplement an income for those still staggering out of the effects of recession. Long commutes and penny-pinching nights in then provide opportunities to show your savings account a little love.
Even with a killer idea, this co-founder is never short of a sales pitch.
Dunayev graduated with a BSc and a BCom (Hons) from the University of Auckland. In 2009 he knocked off an MBA at Stanford University's Graduate School of Business after winning a Fulbright-Platinum Triangle scholarship. It was in Auckland, a year ago where an opportunity arose for Dunayev to channel his talent and passion. Transcribe Me, in a somewhat infant form, was Dunayev’s team’s creation at Auckland’s Start Up Weekend, an event that focuses on building a web or mobile application that could be the foundation of something big. It’s clear the Transcribe Me team discovered its foundation that weekend. But aside from people and growth, vision and determination also plays a part.
“Passion and enthusiasm injects the x-factor into our company. It has to. You’re working for a very long time with very long hours and no pay. You just have to keep on going, based on that passion.”
Co-founder and company director Helga Sonier agrees the team’s chemistry was instant and has only become further refined during the company’s early months.
“Our team [Sonier, Dunayev along with inbound marketing director Chirag Ahuja and operations director Danijel Duvjnak] didn’t know each other prior to Start Up,” Sonier says. “We were all so impressed with each other’s skills, motivation and drive over that weekend. So we immediately committed to moving forward with the project.”
The stakeholders kept their full-time jobs and had Transcribe Me on the side as a part-time project without pay, aside from Dunayev and Igor Feerer, the San Francisco based CFO, who was full-time without a salary from the start. For the former, it’s been “a very up and down experience” – predictably.
“Some days nothing seems to go right. Other days are even worse. You have to have support and drive to keep going. It’s paramount to success and people who really want to see it succeed and prosper.”
Eager to keep the team’s chemistry strong, even beyond country borders, Dunayev wanted passion and commitment to the vision as key themes as the company employed others in the following months.
“Our inner drive and experience was not isolated to the founding team members,” he says. “The investors, the advisors and consultants, literally everyone who has helped us, has shared in our values.”
Dunayev has a ‘no-man-is-an-island’ approach to business and believes any entrepreneurs who believe they are a ‘self- made success’ are fooling themselves.
“It’s essential to have support, which drives you to keep going. It’s paramount to success – people who really want to see your product succeed and prosper.”
Supporters, both financial and academic, who are in sync with the company’s impetus, are valuable when you’re on a growth trajectory. Lessons are learnt quickly and the only constant is change.
Dunayev says the reality of managing a company on a global growth trajectory is like building a plane in mid-flight.
“You’re building a whole lot of things in parallel – the technology, the app, the talent pool of transcribers, the client base, and they all need to come together but they are all so independent.”
Dunayev believes his company is testament to the idea that New Zealand-based high-tech start-ups must be thinking and acting globally from day one.
“That may be the most surprising and counter-intuitive advice I can offer. To target just New Zealand would just spell death for Transcribe Me. I’ve witnessed many other start-ups, from Canada and Israel, and they all start off with a global market in their sights. It’s an opportunity, but it’s also a challenge and you just have to do it. New Zealand is just too limited, not just from a consumer perspective, but for human resources and capital as well,” Dunayev laments.
Although his mentality is overwhelmingly outward-looking, he’s adamant New Zealand is an ideal base and has no intentions of leaving it any time soon.
The stakeholders may not have the most Kiwi names in the book, but Dunayev, Sonier, Duvjnak and their team claim their antipodean origins have presented opportunities along the way.
“New Zealanders have established amazing networks globally. Being a Kiwi, doors open for you that otherwise remain shut for locals.”
The company launched Transcribe Me in Singapore at the DEMO Asia 2012 expo and Dunayev was stoked at the enthusiasm punters had not just for the app but also its country of origin. The day after he landed, he was on national television talking about being an innovative company from an interesting country.
There was, he says, almost instant respect. “I never believed, as foreigners, we would have had huge levels of support. It’s counter- intuitive, but it’s also true in the United States.”
There’s a lot of gold to be mined from Brand New Zealand, and even more so from what Dunayev labels the ‘great entrepreneur ecosystem’, which means support from loads of people – government, investors, even other businesses – at different times. He reckons the company’s early expansion overseas would’ve been dead on arrival without partnerships with organisations such as NZTE, with its trade representatives in Singapore and KEA (the Kiwi Expat Association) with its group in the US.
“In new territory you need loyal partners to make a strong team. I’ve been very fortunate to have great founding team member in Singapore and in California. That was the most important thing in a new country – amazing people on the ground who I trusted, who are deeply passionate about growing the business and who hit the ground running.”
With the the iPhone app now in the Apple store, momentum is building. Dunayev isn’t done, though.
“Our big goal for 2012 is to have near real-time turnaround before the end of the year.”
So, devices that comprehend an entertaining sales pitch or even your own mid-morning rant, with hard-copy evidence? Thanks to some Kiwi high-flyers, that little piece of sci-fi fantasy is just a tiny bit closer to reality.
Top tips for tech startup communication
Communication is challenging for startups that may lack the credibility and resources needed to gain exposure. Internal communications can be just as tricky, with the need to balance differing opinions and personalities. Transcribe Me was no different and found a few key points to overcome the obstacles of growing the profile of a young, unknown company. PR head Helga Sonier tells how.
Embrace every opportunity you get to share your story and your enthusiasm. Although there is undoubtedly a lot on your plate, generating excitement for your idea is contagious. You won’t just attract interest from the media, but also mentors and leaders who are drawn to the team and/or the concept. You want these people!
Remember the team
Internal communications in the early days of startup are critical to ensuring the whole team is steering toward the same objectives. Create forums to hear and measure everyone's ideas. Now is the time to nurture the team dynamic – and don’t forget your fans! Take every opportunity to thank the people who have supported you and others will be keen to follow suit.
Say yes to all offers of exposure, even if it means stepping outside your comfort zone. The publicity can only be good for your business and may afford additional opportunities. Now is the time to let the world, and all of your future customers, know who you are.
A great way of gaining exposure for your company is through competitions promoted and supported by larger companies, often with extensive networks. Many of these also lead to bigger competitions and greater global exposure. Networks are often born and broadened through healthy competition.
If your company’s goal is to be global, then develop international communications from the very start. Set up bases in the countries you want to be (even if it’s just a post office box and a message service). Then endeavor to communicate milestones and developments through local media channels in those places. In other words, meet your customers where they live.
Idealog has been covering the most interesting people, businesses and issues from the fields of innovation, design, technology and urban development for over 12 years. And we're asking for your support so we can keep telling those stories, inspire more entrepreneurs to start their own businesses and keep pushing New Zealand forward. Give over $5 a month and you will not only be supporting New Zealand innovation, but you’ll also receive a print subscription, an Idealog t-shirt and a copy of the new book by David Downs and Dr. Michelle Dickinson, No. 8 Recharged (while stocks last).