Members Only
lock close icon

What Goes Into Making a Voice-Based App for Smart Speakers?

There are a lot of voice-based applications for smart speakers out there, but how are these built?

S Aadeetya
Explainers
Updated:
Amazon and Google have entered India with their artillery of voice-based applications.
i
Amazon and Google have entered India with their artillery of voice-based applications.
(Photo: The Quint/Rahul Gupta)

advertisement

The Indian developer community doesn’t get the credit it deserves, but Amazon is putting this debate to rest once and for all.

The Jeff Bezos-owned technology giant envisages a world built around the Star Trek vision, where humans should be able to do anything through a voice command. And it is already happening.

‘Hey Google’ and ‘Hey Alexa’ will pretty soon become popular trigger words in your circuit/household, and voice will be at the forefront of the same. The advent of voice-commands is slowly getting thrust upon computing platforms, and developers are taking notes.

It’s not as lucrative as mobile apps just yet, though. So, what’s pushing developers in the country to make voice skills for Alexa-like platforms, and how do they even make it? We spoke to multiple developers and Amazon to know the ropes of this business.

What is a Voice Skill?

Making an Alexa skill consists of two parts: you need to build the code for the action that will be performed and host it somewhere (in the cloud mostly); then you need to design the voice skill and plug it with the action code.

That’s basically how the base for a voice skill/action shapes up, according to Rajendra Kadam, an undergrad student who’s made a few skills for Amazon Echo in India.

Voice skills support a variety of coding languages including Java, C++ and Python. If you don’t know any of these, you can still design skills with a bit of effort.
Rajender, a developer making skills for Alexa
(Photo: The Quint/Rahul Gupta)

That effort has somewhat been minimised by Amazon. The company’s got a skill-building interface that merely requires the developer to push his code to Amazon’s Web Service (AWS) server and host it on the cloud.

This exercise, according to Dilip RS, country manager for Amazon Skills in India, has resulted in the creation of about 1,000 skills by developers, with one skill getting added every 90 minutes.

The main focus of creating a voice skill should be about building a good conversation. If the skill can’t do that then the user isn’t going to use the skill. If there’s a skill for flights, asking where you want to fly, it needs to interpret the different ways in which the consumer might ask for its features. 
Dilip RS, Country Manager, Amazon Skills, India,

Popular Use of Voice Skills in India

  • Play music, ambient sound
  • Gayatri mantra or Quran recitals
  • Bollywood trivia
  • Set reminders, alarms
  • Play quiz-based games like Simon Says, True or False

According to Dilip, the skill needs to understand the different accents with which a user might ask a question. This could be anything, from calling out city names in their own way (Bangalore or Bengaluru or even BLR) to pronunciation.

Amazon is hosting varied forms of Alexa Days to connect with developers across the country, and it has even tried out a few virtual hackathons with brands like HackerEarth.

Unlike physical hackathons, with our model, Amazon is able to connect with developers in India from all corners. The scale at which virtual hackathons take place is much larger, and can be spread over months, rather than few days. 
Sachin Gupta, CEO and Co-Founder, HackerEarth

For a voice skill to succeed, Dilip highlights the developer needs to anticipate the users’ actions and what he or she might say should be the trigger word for the developer to use.

Voice Skills vs Mobile Apps: What has Changed?

In comparison to mobile apps that are dependent on visual elements via a user interface (UI) for a better user experience (UX), voice skills are based on voice user interface (VUI), which is based on convenience.

Developers across the world are quickly adapting to the change in trends, much faster than mobile apps, according to few developers we spoke to.

Vishnu Saran, who started a small venture just to make voice skills for Alexa, has made multiple skills for businesses as well consumers. Vishnu, like Rajender, shifted from mobile apps to voice skills because he feels that’s where the future is headed.

(Photo: The Quint/Rahul Gupta)
Unlike mobile apps, where everything is under your control and you hand the user with all the steps of usage, voice skills are all about his (the user’s) actions and what he might say. 
Vishnu Saran, developer for Alexa skills

Dilip, Vishnu and even Rajender are part of a growing ensemble in India that is excited about voice skills and this is only going to increase in the next year or two.

ADVERTISEMENT
ADVERTISEMENT

Where's The Money?

The mobile app revolution started more than a decade back, and for developers it has become a money-minting business. But voice skills are barely a couple of years old as a product, so why are developers rushing into this space and are they getting incentives for their efforts?

Vishnu feels that Indian developers are on their toes when it comes to building skills and believes making money from voice skills won’t be a long wait.

Developers are learning faster, migrating from apps to skills. It’s not going to take long for skills to start making money, not four to five years like mobile apps. Adoption is much faster than many expected.
Vishnu Saran, developer for Alexa skills
Demand for voice skill designers and UI engineers should pick up by 2019 according to experts in the domain.

As Sachin pointed out, HackerEarth has already seen developer traction from cities like Pune, Cochin, Bhopal and Vizag among others. This clearly shows the interest in voice-based applications is reaching nationwide levels.

Over 2,854 developers have registered for various Alexa hackathons, with over 205 ideas for voice skills submitted. 

Even Amazon understands the challenge for developers to work on an ecosystem that doesn’t earn them money. Which is why it offers AWS credit worth $100 (Rs 6,500 approx) to the developer to help them start off without incurring costs on hosting their skills on the server.

In addition to that, developers hosting/publishing their skills get free Alexa merchandise and can be a part of contests to actively participate in the circuit.

Developers are given Echo for free if their skills reach a certain number of users. 
Dilip RS, Country Manager - Amazon Skills, India

Interestingly, in US, Amazon pays developers for good quality skills and it’s quite possible that gradually Indian developers will also get this opportunity.

For now, Dilip and co are focused on getting developers to understand the technique of building voice skills.

Voice is the Future

All said and done, it has become evident that voice skills will become the primary task of developers in the years to come. Developers are building their skills for the the big pay outs, which according to experts is not far away now.

Unlike in the UX world, developers know what voice skills should be built and they are starting of in the domain by building simple, fact-based skills. 
Dilip RS, Country Manager, Amazon Skills, India

Vishnu believes transitioning from visual to voice apps is a learning process.

Google Assistant moving beyond smartphones now. (Photo: The Quint)
We’re heading to an era where communication with devices is becoming natural. Voice-based UI will push developers to further understand design in other ways, giving a hands-free experience to the user. 
Vishnu Saran, developer for Alexa skills

Amazon’s early debut with Alexa in India has given it a first-mover advantage over Google’s Assistant and it is looking to make the best of that. Everyone in the developer circuit is expecting a quicker transition of India's evolved developer base (not very different from its global peers).

Right now, Indians are learning to crawl in the voice-skills domain. Their interest in creating simple, time-friendly skills could catapult them into working on complex voice-based apps in the future and earn them revenue.

(The Quint is now on WhatsApp. To receive handpicked stories on topics you care about, subscribe to our WhatsApp services. Just go to TheQuint.com/WhatsApp and hit Send.)

(At The Quint, we question everything. Play an active role in shaping our journalism by becoming a member today.)

Become a Member to unlock
  • Access to all paywalled content on site
  • Ad-free experience across The Quint
  • Early previews of our Special Projects
Continue

Published: 24 Apr 2018,07:41 AM IST

ADVERTISEMENT
SCROLL FOR NEXT