Jennifer Dulski sold an AI startup to Google in 2009. Now Rising Team's AI trainer, RT, coaches managers and teams. How it works and how it stays private.
Rising Team is an AI leadership and effectiveness personal trainer: an AI coach called RT that learns you and your team, then coaches you, practices hard conversations with you and runs team building sessions. Founder and CEO Jennifer Dulski built it because the executive coaching and training she got as a leader never scaled to the hundreds or thousands of people on her teams. On Episode 52 she showed us how it works and what it changes inside a company.
Jennifer did not start in tech. She began as a high school teacher and founded a nonprofit that helped first-gen students become first-gen college graduates. She says that turned out to be useful later:
a lot of difficult, challenging teams are not that different from difficult, challenging teens, as I have found sometimes.
— Jennifer Dulski, Episode 52
Then came twenty-five years in tech: ten at Yahoo, several each at Google and Facebook, and five helping to run Change.org. As a division leader she had human executive coaches and was sent to great trainings. The problem, she says, was that none of it scaled back to her teams. It was too expensive and HR did not have the bandwidth. In 2020 she spun off and bootstrapped Rising Team to give people at every level that kind of high-touch support without the high-touch cost.
Watch Jennifer Dulski on AI transformation in Episode 52 (starts at 6:43) · Episode 52 notes and transcript
Sam asked whether AI even existed when she started. Jennifer's answer: it existed, it just was not available to everyone. Her first startup, DealMap, was an AI company. In 2009 it used natural language processing and sentiment analysis on tens of millions of local reviews, with models her team built and trained themselves, coding examples late into the night.
Google bought DealMap. Jennifer says those early pieces now sit inside Google Maps, in the short descriptions that tell you a place is casual or good for kids. Her original 2020 outline for Rising Team already said "personalized custom AI coach," but she could not build it the way she wanted then. So Rising Team started as team building software that any leader could run, even remotely, without an outside facilitator. Once the LLMs launched, the team built the AI trainer layer on top.
That order matters. Jennifer says the team building sessions are what let RT understand your team, so its advice is specific instead of generic:
Tom really likes his feedback super straight on. Let me write you a script. Let me practice that conversation with you so it will land with Tom.
— Jennifer Dulski, Episode 52
Jennifer's view is that most companies have missed that it is not enough to throw tools at people. She points to research from this summer by BCG, McKinsey, Gartner and others:
And the BCG data says 70% of whether your AI transformation is successful or not lands on whether the people are able to adapt to these tools and still work with each other.
— Jennifer Dulski, Episode 52
She also teaches at Stanford GSB, and she says the science of high performing teams has not changed: people who give and receive feedback well, know what is expected of them and make others feel valued. What falls short is the old delivery model. Her analogy is a triathlon. Put 30 people in one classroom and teach them the same thing at the same time, and you ignore that some are strong swimmers and some hate to swim, and nobody gets to practice with their own team.
So RT works like a trainer that does an intake and then switches modes based on your goal. It can coach you, run a working session that produces something like a script or a review, role play with you voice to voice, and run team building sessions. The more team building you do, the more it personalizes to you, your team and your company.
Jennifer says most products in her space measure success by whether people use them and like them. People do both with RT, she says, but the question she cared about as an org leader was whether it moves anything else. According to her, customers see statistically significant lifts in manager effectiveness, employee engagement, eNPS, retention and productivity.
Retention is where she puts the ROI. AI has made applying for jobs so easy that an opening can draw 10,000 applicants, but sorting, interviewing and retraining still eat time and money. Rising Team sees retention lifts between 10 and 33%, she says.
The customer list is wide: Airbnb, HP, startups, Blood Cancer United (formerly the Leukemia & Lymphoma Society), K-12 teachers and admins practicing parent-teacher conferences, and banks. Jennifer says the most traction is in frontline teams like bank branches, customer support and IT, the places that are hardest to reach and most under supported. If you follow how AI is changing people work, our AI in recruiting and HR page collects the related episodes.
In the demo, RT starts by asking about your experience, challenges and goals, and which company value you want to work on. It then summarizes what it learned and builds a one-of-one growth plan. Jennifer's own plan covered leading through change, making her team feel valued and speeding up AI adoption, with each track broken into working, coaching and team building sessions.
Like a human trainer, RT finds time on your calendar, schedules the session and sends reminders through messaging. You can also go to it on demand. If you have a hard conversation coming up, the first thing RT does is show you the people on your team. Jennifer says that is the big difference from a general tool like ChatGPT: it already knows your team.
Jennifer argues the annual engagement survey is dying. It runs once or twice a year, takes months to analyze, and nobody knows how to act on it. With RT, leaders see in real time what people are working on, in a rolled-up, anonymized form. Examples she showed included practicing for a promotion denial, handling underperformance, and, at financial institutions, the stress of prepping for licensing exams.
After every session people rate it out of 10. Jennifer says the average is eight and a half and the median is nine across customers. Jordan checked the privacy line: leaders see how many people are asking about a raise, not that a specific person asked. Jennifer confirmed it, and said data can be sliced by level, location or function but never down to the team level.
RT also has built-in guardrails tied to the EU AI Act. It will never help make a hiring, firing or rating decision, and companies can add stricter rules. Financial services customers do, and RT sends people back to HR if a conversation goes somewhere it should not.
In the news segment, Sam said he is using Siri again on iOS 27, and complained that his Nest Hubs still do not have Gemini. Jennifer, an ex-Googler who carries both a Pixel and an iPhone, said Apple took so long that most people have other tools by now. Her bigger point: the best AI features are the ones you do not have to turn on, because they just work while you do something else. We covered more of Apple's AI story in Episode 56 with one of the people who built Siri.
Jordan thinks every assistant is still far from always-on. Then Jennifer asked us a question she is weighing for her own product: is memory helpful or creepy? Jordan's answer was helpful. He wants his phone to notice he has not set his morning alarm and ask whether he wants the usual time.
Setting the alarm is not hard, but remembering to set your alarm is the hard part.
— Jordan Metzner, Episode 52
Jennifer's demo, the full privacy discussion and our iOS 27 debate are all in Episode 52 of Built This Week, with the full transcript on the episode page. You can find more from her on her guest page, and more on how teams are using AI day to day on our AI in productivity page.
