Exclusive: Trevor Project alums debut call center training startup ReflexAI
Illustration: Tiffany Herring/Axios
High-stakes call center training business ReflexAI raised $3.3 million in seed funding, CEO Sam Dorison tells Axios exclusively.
Why it matters: The recent launch of the nation's mental health hotline 988 has spiked call volume to crisis centers, which need resources to manage that increase.
Driving the news: Cities including Washington and San Francisco have seen an estimated 9%-12% of pleas for help go unanswered.
- "There's a lot more volume coming into the system," says Reflex CTO John Callery. "We're here to really help a lot of these organizations meet that need."
Details: Footwork led the round, joined by Emerson Collective, Altman Capital, Gaingels and several individual backers.
- Funds will be used to build Reflex's core operating platform and team and hone examples of use cases for its tech.
- Dorison foresees Reflex raising a Series A sometime in the next two years.
How it works: Reflex provides AI-powered training and quality assurance tools for call centers in industries such as crisis response, 911 dispatch and health care.
- Dorison likens the tools to flight simulators for pilot training, where they can be used both to help train people and also to assess their abilities.
Flashback: Dorison and co-founder Callery met as executives at LGBTQ+ suicide prevention nonprofit The Trevor Project, where they developed Reflex's tech backbone.
- At his former employer, training volunteer counselors was "one of the least scalable aspects of the organization," Callery recalls.
- He and Dorison began experimenting with automating the training's role-playing component using early versions of ChatGPT or GPT 2.
- The pair soon started getting inbound interest from related organizations interested in using the tech to help them scale their own efforts — leading them to create Reflex.
What they’re saying: Given the potential applications of Reflex across mental health, Dorison and Callery stressed the need to make their AI as transparent and explainable as possible, including the evaluation scores they provide to employees.
- There should be "a very clear understanding between the score they got and what they need to do to improve it," says Callery.
- "If you're in the mental health space, you need that level of transparency — you can't just trust an algorithm," he adds.
