"She has a great skill on giving the presentation in terms of stories. She got a completely different picture out to people… That was a fantastic presentation. Keep doing it."
Pavan Kumar
JPMorgan Chase
Case study · NXT AI 2026
Showing the people who build AI what happens when it meets a real business.
At NXT AI 2026 in Boston, Galina gave a room of AI researchers, academics and enterprise technologists something they rarely get: firsthand evidence from the operators actually using their technology, and a practical path to adoption that works with people instead of around them.
- Event
- NXT AI 2026, hosted by CO-IN Symposia · Boston, November 2 to 4, 2026
- Talk
- The Reality Behind AI Adoption (20 minutes)
- Audience
- AI researchers, university faculty, PhD researchers, engineers and enterprise technology leaders
The challenge
The builders rarely hear from the users.
The public conversation is split between two loud stories: AI is coming for your job, or AI agents will do everything end to end. Neither matches what businesses are actually experiencing.
The NXT AI audience sits on the building side of that gap. Researchers, faculty and enterprise technologists shape the tools, the curriculum and the internal rollouts. They needed a grounded view of where adoption really stands, why it stalls, and what the next generation of products should solve.
The approach
You're not behind. You're early.
Galina built the talk on her Artificial Reality interviews with business owners and operators, asking each what's working, what's broken and what still needs to be built. According to the U.S. Census Bureau, only about one in five U.S. businesses use AI to produce their goods or services.
- 01What's working: A civil engineer used Claude to turn a day of structural calculations into minutes. A capital markets team cut training video production from weeks to hours with Synthesia. A health coach built a client intake system on 15 Claude skills.
- 02What's broken: ChatGPT reviewed a 45 page lease for a first time restaurant owner, but never told her she needed a mechanical, electrical and plumbing engineer. That gap cost her a month and a $30,000 contractor. An AI drawing checker flagged thousands of issues and made more work, not less.
- 03What still needs to be built: Operators in unrelated industries asked for the same four products: client memory, a verification and trust layer, natural voice intake, and one pipeline across AI tools.
The plan
The A.D.A.P.T. framework.
Paired with a workforce strategy: promote from within. Prosci's research shows projects with excellent change management are about 7 times more likely to meet their objectives.
- AAssess the workflow. The real one, not the diagram.
- DDefine one problem. Prove it, then expand.
- AActivate your team. Peers build, peers train.
- PPreserve human judgment. Decide what stays human before something goes wrong.
- TTrack the results. Measure before and after.
AI adoption doesn't happen TO you. It happens WITH you.
The audience response
A realistic, human picture of AI adoption.
Attendees from banking, academia and AI research said the same thing in different words: the talk replaced fear with something real.
"What I like best about your presentation is that you were very clear, you speak very nicely, you're very charming, and on top of all that, I really like the way that you present that you have to reach people where they are at and help people with their problems and take that user-centered approach. And that's a really great message that really resonated with me."
Matt Wright
Professor and Chair of Cybersecurity, Rochester Institute of Technology
"In this era of all of the discussions being polarized… I felt like her talk was really real. It was realistic. No big words, very friendly, realistic talk. And I think we really need that."
Darya Ramezani
PhD researcher in AI and society, Drexel University
The takeaway
Even the people who build AI needed real evidence and a plan built around people.
A banker, a cybersecurity department chair and a PhD researcher each named the same strengths: storytelling, a user-centered message, and a realistic tone in a polarized conversation.
In the room
The Reality Behind AI Adoption, delivered at NXT AI 2026 in Boston.