How We Built a Chatbot for Real People

Original

The App ecosystem is broken. There are millions of apps out there, and it is hard to find that one that suits you. Platforms do not seem to be investing enough in improving their algorithms (or even applying AI at all) to make them more discoverable or relevant to individual users.

For users, the experience with apps has become laborious: you open the app store, search for an app, scroll to find what you want, download it, install it, open it, sign up, go through a tutorial and voila! After a lot of clicks, you finally get to the real deal.

For developers, it’s extremely expensive to get their app discovered, to retain users and to monetise them. On top of building a truly viral app, to succeed in this environment, it takes serious investment in marketing, PR and community.

And then comes a new hope: chatbots. Earlier this year, Facebook, Slack and Skype announced they’re open for (chatbot) business. They’ve released developer tools and frameworks to encourage more businesses to create bots that currently range from giving you the weather forecast, assistance in finding places to eat, curating news for you, etc.

However, if you talk about bots with Joe Bloggs, you’ll hear “chat-what? How does it work? Where can I find them?”.

Welcome to the world of new platforms. That’s just a start for sure - well, a start that’s been happening for a while. One of the first chatterbots (as they were called back in the 60s) was Eliza, a bot created at the MIT Artificial Intelligence Lab. Eliza never became widely adopted. Why not? Perhaps it’s because it was a mere experimentation, never intended to become mainstream, or even that AI, NLP and machine learning weren’t actually as evolved as they are nowadays, making Eliza not so believable. Or perhaps it’s simply because we humans were just not ready for interacting with bots.

Let’s not forget how far we have come since the 60s: as of 2016, over 2.5 billion people have at least one instant messaging (IM) platform installed on their phones, and research suggests this will increase to 3.6 billion globally within the next two years. That’s half of the world’s population with an IM on their mobiles.

According to this year’s Mary Meeker Internet Report, messaging apps are becoming our second home screen. The average mobile user only uses 12 apps on a daily basis, with 80% of their time spent on only 3 of them (Facebook, Whatsapp and Chrome).

The world of apps and chatbots will go through an exponential transformation in the next few years, thanks to the pervasive use of IM.

We at BOLDR are on a mission to democratise coaching. We help people learn more about themselves and improve life and work skills via a chatbot that lives on their IM. It’s like having your very own (AI-powered) virtual coach in your pocket.

Here are the top 5 things we learned while building BOLDR that I’d love to share with you:

1. Start with the goal in mind

Whenever a new platform arises, a lot of people jump in to try it out, and create the most gimmicky uses of it. Tech for the sake of tech. This new gen of chatbots is no different. AI has shown fantastic advancements over the last 2 years, and Facebook, Skype and Slack now offer a platform for developers to build on. That’s all great news - but wait, not so fast.

Thinking about the real problem you are trying to solve is key if you want real people to use your service. Don’t build a chatbot just because, or you risk being like the early apps with no use to humanity, and, even worse, irrelevant and impossible-to-monetise. 

Find a problem someone truly has and work on the best solution. Would a conversational UI be the best way to solve this problem? Make a chatbot.

2. Frameworks are temporary solutions

When my co-founders and I started BOLDR, we chose to go with the Microsoft framework. This helped us get our bot off the ground fast. If that’s your aim, go for it. Do it quick and dirty. As your chatbot starts showing signs it could take off (i.e. when relevant users give you positive feedback), spend proper time to think of your tech stack, the engine you’re building and the value you’re bringing to users. How will you create valuable IP to your company?

3. Deep learning, machine learning and NLP are the cherry on the cake at early stages

Since everything AI has the potential to make your bot super smart, you will be tempted to invest on those first. Don’t. Or at least, think twice. Is AI crucial for your chatbot to give value to your users? If not, find another way to create and validate your MVP.

Bots are conversations, and conversations are very personal, so take the time to learn everything about how your target audience interact with the service you want to provide. Then spend a lot of time perfecting it. Your aim is to use the latest advancements of tech to create a service that’s 10x better than what other companies give to your target customers.

4. Find the perfect advisor in your niche

How will your chatbot help your users? Will it help them schedule a doctor’s appointment or connect them with the right lawyer? No more weather please - I get it, it’s fun, but Google can do this pretty well ;)

Once you have defined your niche, go find advisors who are highly experienced in the field. They’ll help you tailor the ideal experience and save you a huge amount of time by focusing on what really matters to your audience.

5. Build the best team to make it happen

It doesn’t matter if you’re a non-techie exploring a new concept or venture in this area, or a developer who’s keen to learn and create something useful and out of the ordinary.

If you’re serious about making a business out of it, gather the right team to help you get there. Great services, bots and apps (that are truly marketable) are created by amazingly talented multi-disciplinary teams, not an individual.

Are you also building a chatbot? I’d love to hear your comments on whether these tips are useful and exchange notes with you. Go on, tell me more.

Roberta Lucca is Co-founder and CEO of BOLDR, a chatbot that aims to help people become more focused, confident and better versions of themselves. Their vision is to democratise business coaching, a market valued at $11billion. You can sign up for the beta version at boldr.me or reach Roberta on Twitter @olicca.

Would you like to be our next guest author? Find out how here!

This is a guest blog and may not represent the views of RE•WORK. As a result some opinions may even go against the views of RE•WORK but are posted in order to encourage debate and well-rounded knowledge sharing, and to allow alternate views to be presented to the RE•WORK community.

Learn more about the future of chatbots at the next Virtual Assistant Summit, in San Francisco on 26-27 January. Speakers include Roberto Pieraccini, Director of Advanced Conversational Technologies at Jibo; Jordi Torras, CEO & Founder of Inbenta; Rachael Tatman, PhD Candidate at the University of Washington; and Alonso Martinez, Technical Director, at Pixar.

Discounted tickets are available until 7 October, book your pass at an Early Bird rate and view more speakers on the event website here. View all upcoming RE•WORK events here.

[Image via www.fastcocreate.com]

Guest Blogs A I Mobile Devices Chatbots AI Assistants NLP Deep Learning Algorithms

0 Comments

Search

Recommended Posts

Latest Posts

Upcoming Events

Deep Learning Summit London

21 September 2017, London

The Deep Learning Summit is at the forefront of AI. Explore the impact of image & speech recognition as a disruptive trend in business and industry. How can multiple levels of representation and abstraction help to make sense of data such as images, sound, and text. Hear the latest insights and technology advancements from industry leaders, startups and researchers.

AI Assistant Summit London

21 September 2017, London

The next generation in predictive intelligence. Anticipating user & business needs to alert & advise logical steps to increase efficiency. The summit will showcase the opportunities of advancing trends in AI Assistants & their impact on business & society. What impact will predictive intelligence have on business efficiency & personal organization?

Track 1: Deep Learning Summit Montreal

10 October 2017, Montreal

What are the latest advancements in deep learning research? Where are the most recent scientific breakthroughs? Hear the latest research news from global pioneers in Natural Language Processing, GANs, Reinforcement Learning, CNNs and Unsupervised Learning at the summit.

Connect

Be Sociable

  • Twitter
  • Facebook
  • Linkedin
  • Youtube
  • Flickr
  • Lanyrd
  • Instagram
  • Google plus
  • Medium