Enable natural interaction with your app using a conversational interface powered by the Watson Developer Cloud APIs.

There's a new direction coming for user interfaces, and we want you to help shape it. Pages of forms and static information are a thing of the past: why not let your applications talk to your users? IBM Watson has four new services that make it simple to set up a conversational interface to interact with your users as if they were talking to a person – Natural Language Classifier, Dialog, Retrieve and Rank, and Document Conversion.

We want to see you use this technology to build a conversational application. It could be a chat bot, it could be an automated phone system, it could be something no one has ever seen! Take an existing process and make it more human (but without the human). An advisor to walk you through filling out your taxes, maybe. Or an assistant to help you with a tricky repair hands-free. We want to see your best ideas.

What will you build with Watson?

Important Dates:

  • Start: February 29 at 8am PST
  • End: April 15 at 11:45pm PST
  • Time Zone: Pacific Standard Time

View full rules



  • Countries: United States only (including District of Columbia)
  • Participants: Individuals (over 18 years in age); Teams of no more than 5 people; Small organizations (less than 50 employees); Must not be employees of, or have any affiliation with IBM
  • Entries: Limit one (1) entry per person/team/small organization



  • Use two or more of the following Watson Developer Cloud APIs: Dialog, Retrieve and Rank, Natural Language Classifier and Document Conversion. Feel free to use as many additional APIs as desired.
  • The application should be able to interact with users using natural language sentences; don't rely solely on multiple-choice or yes/no questions
  • Ability for text and/or voice input. Use of additional Watson services, such as Speech to Text and Text to Speech, will be looked upon favorably.
  • Include publicly-available code and training data in addition to your working demo 
  • The use of any and all coding languages is permitted, including Python, Node, etc.
  • Registration is not limited and interested participants can register at any time throughout the course of the hackathon.
  • Team size can range from 1-5 but cannot exceed 5 people.
  • Your application can be hosted on any hosting platform, including Bluemix.
  • Participants who leverage Bluemix and have reached their 30-day trial limit will receive a free extension for the duration of the competition.

How to enter

  1. Register for the Challenge on the DevPost site. To complete registration, users will be asked to create a submission page to host their final application.
  2. Create a Bluemix account (if you don’t already have one) to access all Watson Developer Cloud services.
  3. Participate in the Challenge by creating a working web app that converses with users and contains two or more of the following WDC APIs: Dialog, Retrieve and Rank, Natural Language Classifier and Document Conversion
  4. Host the application on a hosting service, such as Bluemix.
  5. Make app code publicly available on a site, such as GitHub.
  6. Submit project via official contest submission page on DevPost
  7. Make sure your submission includes all of the requirements


Dan O'Connor

Dan O'Connor
Senior Software Engineer / IBM Watson

Charles Quincy

Charles Quincy
Business Development Manager / IBM Watson

Judging Criteria

  • Best Use of Watson Services
    The judges will be looking at how well the participant used a combination of Watson Developer Cloud services to enhance their application.
  • Relevance to the Challenge
    Each qualified submission will be judged based on the extent to which the application solves a real-world problem using a conversational interface.
  • Quality of the Code
    Your apps will be shared with our developer community, and to make sure others can follow what you have done the code should be easy to read and reuse.
  • Overall Appeal
    From a user standpoint, how easy is the application to use, how well could the application function when applied in a real-world setting.