AI Product Leader
Travis
Hollingsworth
I partner with founders & investors to take AI products from zero to one.
Founder, CPO, investor.
Previously led frog Ventures in San Francisco.
Stanford GSB Sloan Fellow.

Research
ProductAI Product Design Is Relationship Design
What does it mean to design a relationship with a machine that can think, act and evolve?
ThesisWe have traditionally designed products as objects: bounded, predictable things that people learn to use. But what happens when the object begins to learn about the person using it? AI transforms product design into the design of a two-way relationship, shaped by trust, agency and mutual adaptation.
SaaSThe SaaSpocalypse
What if the real value of software was never the code?
ThesisThe prediction that AI will kill SaaS rests on a peculiar misunderstanding: that software is simply code. The best products embody years of accumulated intelligence—learning from thousands of users, resolving exceptional cases, and distilling complexity into something remarkably simple. AI may make software infinitely easier to produce without making great products any easier to conceive.
Venture CapitalThe Mental Models of Venture Capital
How does an investor's understanding of their own work shape what they want from AI?
ThesisSome venture capitalists see their work as finding deals. Others see it as cultivating relationships. Research across 15 international funds revealed four distinct mental models of investing, each pointing to different possibilities for AI. But the most compelling opportunity may lie beyond deal screening and diligence: helping investors discover exceptional people, cultivate relationships and recognise tomorrow's founders long before a company exists.
Gen ZComing of Age with AI
How does a generation growing up alongside intelligent machines decide what to trust, learn and delegate?
ThesisYoung people are developing an entirely new relationship with intelligence: something they can turn to for answers, guidance, encouragement and practice. Yet unlike older generations, many seem remarkably unimpressed by generic AI. Drawing on design research with 16–24-year-olds across the UK, US and Italy, we explore the mental models shaping their expectations of AI—and the tension between wanting assistance and developing their own confidence, judgment and independence.
LearningBeyond Agents & Efficiency
What if AI's greatest opportunity is not doing more for us, but making us more capable?
ThesisThe AI industry is preoccupied with automating human work, but efficiency is only half the opportunity. Drawing on design research with young people through Sky and Prosper, we explore AI that develops human capability through realistic practice, personalised feedback and reflection—rather than simply completing tasks on our behalf. The real measure of success is not what the machine accomplishes, but what the person becomes capable of doing without it.
AI MoatsMoats: What Can AI Applications Still Own?
When frontier models can do almost everything, where should founders invest in lasting differentiation?
ThesisEvery advance in frontier models threatens to erase yesterday's competitive advantage. But the most valuable applications accumulate what general-purpose models cannot simply inherit: privileged context, trust and responsibility for real-world outcomes, and the ability to learn from experience. These advantages compound through use, potentially making an application more valuable even as the intelligence beneath it becomes ubiquitous.
VenturesThe AI-Native Playbook
If a tiny team can build almost anything, what does it take to build something people actually need?
ThesisAI makes it possible to build in days what once took months, but speed creates a new illusion: a working prototype is no longer convincing evidence of product-market fit. Drawing on experience developing AI-native ventures, we explore a different playbook—small teams working closely with domain experts, designing guardrails for unpredictable behaviour and continuously learning from users, evaluations and real-world outcomes. The real advantage is not how quickly a company builds, but how quickly it discovers what works.
What can AI applications still own?
Three advantages that compound through use.
Attunement
It knows what
matters to you.Privileged context
Autonomy
It earns the right
to do more.Earned responsibility
Adaptation
It learns what
actually works.Learning from experience
Builds with every use
Products
AI companions, copilots and new product directions. My work spans early research, product strategy, prototyping and leading teams through development.
| Product | Insight | Role | |
|---|---|---|---|
| SonarAI companion for teens | Teens value a space to be honest without managing someone else’s feelings. | Founding CPO Scaling across US schools | |
| TuneddAI copilot for venture investors | VCs want help building relationships with great founders, not automating their decisions. | Founding CPO Live pilots with top-tier VCs | |
| MetaAI personas in social media | Empathy is experienced within the structural approach of an AI conversation. | Product Team Lead Rolled out to 4BN users | |
| P | ProsperAI simulations for college | Confidence is built through private, personalized practice with encouraging feedback. | Founding CPO Public launch in 2027 |
| CetipFintech data for credit decisions | Sharing asset data reduces systemic risk and helps spot fraud. | Founding CPO Acquired for $3.9BN | |
| Apple & HoagPreventative primary care | Apple Watch feedback and health goals help turn concerns into actions. | Product Team Lead Launched with Apple Watch | |
| SkyInteractive AI media for Gen Z | Young professionals value insights from those ahead of them, and AI personas can replicate aspects of mentorship. | Product Team Lead Public launch in 2027 | |
| McGraw-HillTeacher copilot | Educators need to adjust lesson plans overnight to reflect evolving student progress. | Product Team Lead Partnered with Claude for Teachers | |
| EthixbaseAI agents for legal & compliance workflows | The biggest barrier to agentic delegation is nurturing trust, which is earned through consistency. | CPO Agents launching in 2027 | |
| EverwayCopilot for Special Education | Educators need tooling to tailor learning content to match their students' personal skill levels. | CPO Acquired TeachTown in 2026 | |
| Brainpool AIContext layer for private LLMs | IP-rich businesses need bespoke & safe AI models. | Advisor 300%+ enterprise growth | |
| lululemonBody avatars for online shopping | Fit is a sense that comes from recognizing your body & imagining feeling via movement. | Product Team Lead Partnership with Epic Games | |
| Lucid MotorsLuxury electric vehicles | People want the experience of a luxury car without wanting the commitment of ownership. | Product Team Lead Robotaxis with Uber & Nuro | |
| Mammoth BiosciencesGenetic testing at home | Testing at home while unwell can be daunting and lonely. | Product Team Lead $1.4BN valuation | |
| Sage TherapeuticsPersonalised support for people navigating depression | AI helps privately organise thoughts before speaking with a clinician to prepare & build confidence. | Product Team Lead FDA Fast Track | |
| AllioAI portfolio tools for investors | People think about what their money is for before they think about how to invest it. | Product Team Lead New MCP layer for advisors | |
| MasterGrassVideo platform for Gen Z | Young people gain gardening confidence by observing others and learning contextually. | Product Team Lead Funded by Walmart & Home Depot | |
| Gates FoundationShared intelligence & agents for vaccine research | Researchers are reinventing the wheel each time, and AI can dramatically shrink timescales. | Advisor Launching in 2027 |
Investors
Investors I’ve worked with across venture capital, growth equity and private equity.
Talks
Ideas worth questioning, arguments worth having, and conversations that might change how we see the world.
What are AI Product Moats?
What makes a product irreplaceable? Attunement, autonomy and adaptation.
Should we welcome the end of jobs?
A debate about freedom from labour and a life beyond employment.
The AI Product Playbook
We used to build predictable products. Now systems think, act and grow.
Changing Our Minds
What have we got wrong about AI—and which convictions still hold?
What Does It Really Mean to Be a Founder?
The hidden doubts and pivotal choices behind building a company.
Our Personal Companion of the Future
When does a useful AI assistant become something we confide in?
Unlocking Meaningful Conversations
Getting beyond small talk to the conversations that change us.
How to Disrupt Yourself: Life in the Entrepreneurial Home
When home becomes work and life becomes a project, what remains?
What Does It Take to Build Something New?
Turning an idea into something real means confronting uncertainty.
Press
Bio
I’m interested in the relationships we build—with one another, with technology, and increasingly with AI.
I’m a product leader, former founder and investor based in London. I work with founding teams and investors to understand users, develop new AI products, and make thoughtful decisions in conditions of uncertainty.
I led frog Ventures in San Francisco, working across AI, fintech, health and other emerging fields. Before that, I founded Norn, a venture built around meaningful conversation and human connection.
My approach combines human-centred design with the experience of building and investing. Research is a way to question assumptions, make choices explicit, and help a team find a direction it believes in.
I studied at Stanford Graduate School of Business as a Sloan Fellow, earning an M.S. in Management, and at Georgetown University. Earlier in my career, I worked at Big Society Capital, Oliver Wyman and Goldman Sachs.





