Skip to content

We Need to Talk About Helping Kids

Why are we building conversational AI systems?

AI for teen mental health support

Yes, our business is about engaging customers through conversations and assessing and addressing their emotional needs.  Yes, that’s important for improving outcomes of sales and customer service interactions.  And yes, these systems can save money and increase revenue.  


But that’s not why we are building these systems.


Our patent pending approach will make companies money so that we can test, train, and fund a 3 layered recurrent neural network system that is designed to have high emotional intelligence and engage in strategic conversations.  The ultimate application of this technology is not sales related, it’s therapeutic.  


We want to build a free therapist that lives on every phone of every teenager and young adult in America.  There are already some great early endeavors towards this goal (most recognizable among them is probably woebot, but there are a few others) but we are approaching this with a more dynamic linguistic technology.  What can do is dramatically increase the abilities of these systems and add important skills to what they are already building.

Sad woman with depression sitting on the floor


We know that early access to coaching and intervention helps prevent small problems from getting bigger.  We know that kids today are under stresses and facing issues that previous generations haven’t dealt with.  We also know they have potential resources that aren’t being fully utilized.  The phones in their pockets don’t need to just be a source of instagram highlight reels and number of likes and follows to compare and compete with other students.  The phone in their pocket can give them access to deeper insights, suggestions, and tools to help them cope with the situations in which they find themselves.


Our approach is a bold one.  The key to making Cyrano work is access to training data that can teach an AI system how to respond.  We believe that using clinical outcomes is too messy and corruptible as a dataset to be viable in a machine learning environment.  Sales of big ticket items and customer service provide the perfect opportunities to have a high volume of conversations around emotionally charged topics.  As we add new industries and companies, we add variety to the topics and conversations the AI system understands.  Once we reach a critical mass of conversations, we can shift the domain knowledge weighting and gear the conversations towards helping kids with their feelings and challenges.  The beauty of this plan is that it self funds once we reach deployment, as there really is no way to get the data we need without being a viable business that provides immediate value for our customers.


This is not a response to recent tragedies.  On both a national and local level we have seen heartbreaking examples of teenagers that need help, and those remind us why what we are doing is worth the effort.  We’ve been working on Cyrano for two years.  My work in my private practice and our work in drug rehabilitation centers reminds us that there are too many kids who didn’t get help they needed until it was too late.  For whatever reason, those kids slipped through the cracks.  It’s an incredibly complicated and sad situation that can’t be solved with a single approach or initiative.  Which is why more people need to help where they can.


We have a vision, a goal, and a plan that can put us in a position to help more kids respond to the significant and trivial problems in their lives in healthier ways.  There’s a long way to go on the journey, but two years in we are moving faster than ever and making more progress every day. wants to be a small part of a solution to a huge problem.


The best chess player on earth is a computer.  We are making a small step towards the goal of having the best therapist on earth be accessible to anyone looking for help.

Sign Up for Our Newsletter