AI Chatbot
Last updated
Was this helpful?
Last updated
Was this helpful?
We built this project at the Washington DC startup weekend in 2017. Startup Weekend is an event where you bring your idea to a real product in 54 hours. I brought the idea - an AI chatbot that can help millions of immigrants access legal service and get legal advice to the event. I pitched the idea and hired a team to turn the idea into reality together. Thanks, everyone on the team for going on this journey with me! Without further ado, let's talk about the story behind the journey.
I had a lot of fun generating the idea. The day when I signed up for this startup weekend, I started brainstorming session. When I was talking to my friend regarding her education opportunities one day, I stumbled upon my first idea - an AI system to improve the current education system. That friend struggled to get enough time to go to school and participate in all the coding education events that I invited her to go to because she works full time and needs to take care of her two kids after work. I asked myself - can technology help those full-time moms to build a customized and flexible education system for them and give them the support they need along the way? Then I started building a real product road map from an abstract concept. As I built up a product road map, I realized that the minimum viable product (MVP) to take on the whole education system is not something that I can build in a weekend. I killed the idea for this event.
The story came to a turning point here. I started doing research again. I knew I want to build some AI products that can solve some really important issues for people and make the world a better place. I looked into those hot topics and trending topics online. Immigration was one of the hottest topics at the time. When I watched the news about immigrants and their children were detained in the airport and waited for ACLU lawyers, I asked myself - why not build an AI product that can help millions of immigrants access legal service and get legal advice easily and at lower or no cost. The idea - an AI chatbot that can help millions of immigrants access legal service and get legal advice was born.
The first thing I did after I locked down the idea is to research about customer demand and competitors. I did some customer qualitative analysis and I knew there was a huge demand for accessible legal service. On the other hand, I did a couple of rounds of competitive analysis research. The only competitor that I could find at the time is a Russian company. They built an AI chatbot that can answer simple questions regarding how to apply US travel, business, and student visas. I couldn't try out the product because they haven't launched their official product yet at the time I found them. In addition to analyzing that Russian company's product, I talked to all kinds of chatbots in the world as part of the research to find out their strengths and weaknesses.
My product idea can totally differentiate itself from this Russian product. For example, my product is not just a bot to answer simple visa questions. It can answer all kinds of common immigration legal questions and can also connect customers to legal service agencies and representatives when necessary. Well, you might ask what prevent them from going after you and doing the same thing and as well as providing similar service. The solution to this is to build up 1) the best AI algorithm 2) the best customer community 3) the best relationship with legal agencies and representatives 4) speed to market 5) dominated brand name.
I arrived at the Washington DC startup weekend event right before it was kicked off. When I walked into the room, I saw a full room of energetic entrepreneurs. You can feel the energies in the room. I knew it was going to be a fun and unforgettable event. After we played the Icebreakers game, the pitch session started. I waited for that couple of people to finish their ideas. I went on the stage. There were only 60 seconds. I wanted to hit a couple of important points - why I came up with the idea, what product is, what competitors we have, what market share looked like(potential profit margin), and how we can build it. I talked about those points in 60 seconds and I knew that I did a great job overall but there was one issue that I didn't leave enough time talking about what team members that I want to hire for this idea.
After the pitch session was done, it was voting time. Everyone was asked to vote for the top 3 ideas they love the most. After the voting process was done, organizers announced the top 10 most voted ideas among 80 ideas. My idea ranked the top 2 among those top 10 ideas! Thanks, everyone for voting for my idea! Then it was time to sell ideas and hire team members. If I couldn't find more than 3 people who want to work on my idea in 20 mins, this idea would be killed. I talked and hired 4 team members - 2 full-time entrepreneurs, 1 lawyer, 1 human rights activist who loves to work with me to bring this idea to reality! So proud of you all! We talked about the motivation behind the idea and what we were going to do in the next two days as well as diving tasks among all of us.
At first, I only wanted to build a chatbot that can answer legal immigration questions and connect to the legal agency when necessary. After talked to the team and startup mentors regarding product design, legal issues, pricing model, relationship with legal agencies, we pivoted the idea from a chatbot to an AI legal forum which not only has a smart AI system to automatically answer common legal questions with disclaimers but also certified legal experts to answer complex legal questions which can't be answered by AI. We have a voting system in the forum as well so that we can help customers identify qualified legal experts and connect with them. Our AI backend learns from human legal experts over time, it can gradually answer more and more complex legal questions. You might wonder what our business model for this product is. We pivoted our business model for a couple of times. We ended up with a licensing model. The team really wants to make this business a non-profit business or at least B corp (social enterprise). That's why serving non-profit legal organizations such as the American Civil Liberties Union (ACLU) with a licensing model is the best model for us.
This is the wireframe that we draw for the product.
We also tried to do user search and marketing during this 54 hours journey. We heard that there was a march regarding the immigration issue going on in front of the White House. Two team members went to the White House to gather user research regarding where immigrants find legal information and what methods they prefer. In meanwhile, I sent out a survey on my social media platform and got a lot of responses back. The most exciting thing among all our research and marketing efforts was the American Civil Liberties Union (ACLU) cold call. One team member made a cold call to one director of ACLU and asked if they would be interested in our product. We didn't get anything back at first and thought they wouldn't be interested in us at all. It turned out we were wrong. They replied back and said - YES! We would love to use your product!
After exhausting work, it was time to present our final products in front of the judges. Our team did an awesome presentation and I couldn't be more proud of all of them! Although it was disappointed that we didn't win, we brought a great idea to reality in 2 days and experienced ups and downs together as well as learned a lot of lessons for our future adventures. Couple startup mentors came and found us and told us that they love the products and don't get discouraged by the judging results and asked us to keep doing it because they all believe that this product is needed! From bottom of my heart, I want to say thanks to my team, the mentors, and all the great entrepreneurs in the events!!
