Before 2017
Built four companies, sold two, then ran innovation inside ThoughtWorks Australia and Sportsbet.
ProfileLeslie Barry
BasedMelbourne, Australia
RoleIndependent advisor · Founder, Exponentially
ReJudgment before implementation
I help enterprise leaders decide which AI bets deserve another dollar. My work sits before implementation: map the portfolio, test the assumptions, gather behavioural evidence, and make a defensible call on what to fund, fix, or kill.
I sell no build and no software. Your team or chosen delivery partner builds the winners. That independence matters because my recommendation cannot become my implementation revenue.
4,000+experiments run with client teams
$68M+generated and saved for clients
50+enterprise teams
2017Exponentially founded in Melbourne
Operating position
Independent by design
Most AI advice arrives attached to a roadmap, platform, or delivery team. Mine ends with a decision. I work for the executive answerable for the result, not for the vendor proposing the build.
The work is practical: compare competing bets on one page, identify what must be true for each to pay off, and get customer or employee behaviour into the decision before the budget is committed.
The record
I have sat on both sides of the funding decision
I built companies, ran innovation inside large organisations, and then founded Exponentially. That operating history is why I care less about how persuasive an idea sounds and more about what it can prove before the expensive part begins.
2017
Founded Exponentially to help enterprise teams test whether ideas deserve capital before they build.
2020 to 2024
Worked with 50+ enterprise teams and trained more than 1,000 practitioners across 200+ workshops.
2021 onward
Guest lectured in the Stanford course where Pretotyping is taught.
Now
Advise executives on AI portfolios, applying behavioural evidence to every fund, fix, or kill decision.
Pretotyping
The method behind my work was created at Google by Alberto Savoia and is taught at Stanford. I trained directly with Alberto and guest lecture in the Stanford course where he teaches it.
Governance
I sit on the PHORIA board and trained with the Australian Institute of Company Directors. The goal is not only a better experiment. It is a decision an executive can defend.
How I work
Evidence first. Verdict second. Build last.
The sequence is deliberately simple. It keeps the cost of being wrong low and makes stopping a normal capital-allocation decision rather than a political failure.
Map the bets
Put every live and proposed AI investment on one page, with an owner, expected value, and next funding decision.
Expose the assumptions
Separate what the organisation knows from what it hopes, especially around adoption, behaviour, and measurable value.
Test before build
Run the smallest behavioural experiment that can change the funding decision while being wrong is still cheap.
Issue the verdict
Fund, fix, or kill each bet, with the evidence attached and the next threshold agreed in advance.
Public record
Follow the evidence, not the biography
These are the pages I would use to check the claims above: client outcomes, public sessions, third-party profiles, and material you can watch or use.
Client and partner evidence
AGL
1,000+ tests a year and $7.5M+ saved after Leslie trained the experimentation team.
Tabcorp
130+ real-world experiments, $12M in avoided costs, and $7.3M in revenue generated.
PEXA
50 pretotypes in nine months, with the method embedded beyond product development.
RACQ
A multimillion-dollar mobility idea killed in three days after real customer evidence showed little demand.
Teaching, speaking, and third-party record
Stanford guest lectures
Four recorded MS&E 265 guest lectures from 2021 to 2024 on quick prototyping, pretotyping, and rapid experimentation.
Speaking, podcasts, and press
A public archive of talks, webinars, interviews, podcasts, and press coverage from 2016 onward.
The Corporate Innovator
A 35-minute third-party podcast on Pretotyping, evidence, risk, and enterprise adoption.
Pretotyping expert listing
Pretotyping.org lists Leslie as the Asia-Pacific expert for keynotes, workshops, coaching, and custom programs.
“Leslie has a deeply purposeful connection to the pretotyping philosophy, and he has done amazing work in industrialising the training, tools and methodologies to truly bring pretotyping to life.”
“The students love the class. They applied pretotyping to test their assumptions before designing their MVPs.”
Watch and learn
See the method, not just the claims
Leslie’s free Pretotyping course links each lesson to its YouTube video, with practical examples and exercises.
Selected writing
The ideas behind the work
How to prioritize AI projects
A five-part framework for ranking value, evidence, adoption, risk, and time-to-impact.
What is an AI readiness assessment?
Why decision readiness matters as much as data, capability, and infrastructure.
What is the AI project failure rate?
A sourced view of the evidence, and what leaders can do before implementation.
How to measure AI ROI
A practical approach to trustworthy inputs and pre-build evidence.
Elsewhere: LinkedIn · The Experimenter's Edge · YouTube
Next step
Bring me the AI bet you are least sure about
In a short call, you will get my honest read on what it needs to prove and whether an AI Bets Audit is a useful next step. No implementation pitch.