For AI founders and startup teams
You're building AI to help people, communities or society. Make sure it does.
I help you see the blind spots and protect what you're building.
Why founders bring me in
"Help me see what I can't."
You're close to the product. The social impact blind spots are the ones you're too close to see.
A social impact review before it ships
"Help me keep it."
Money and growth bring pressure to trade the mission away. Set the guardrails now, while they're still yours to set.
Social impact principles now, then advisory as you grow
"Help me prove it."
Your buyers and investors need to see the return, financial and human, in terms they recognize.
A social impact review with the business case built in, or ongoing advisory
Who I work with
Founders and startups who are human-first, committed to scaling social change, and ready to make the choices that show it.
Some AI choices can't be coded or computed.
They aren't engineering problems an algorithm can solve. They're choices about how your AI impacts people, communities and society, and someone has to make them on purpose.
What it's optimized for
The metric it's tuned to, for whom, and who chose that. Where profit and people pull apart, you decide in the open.
What data goes in, and who owns what comes out
Whether it gives people more control or less, and how you keep it from being abused.
Who gains and who pays
Across race, class, gender, disability and caregiving, and where they overlap.
Those choices decide who your AI serves, who it leaves behind, and whether it supports people or exploits them for profit.
Your engineers can make it work. I help you make the choices that take human judgment, not just data.
Be ready for the questions socially conscious buyers ask
They want to know it pays off, for them and for the people it touches. I help you answer with your product's financial and human return, in terms they recognize.
Will it pay off?
39%
of companies report any impact on company-wide profit from AI. Most of those put it under 5%.
Will people trust it?
50%
of US adults are more concerned than excited about AI, up from 37% in 2021.
Three ways to work together
Social impact principles for your product
Human-first principles for how your product treats the people it's for and the people it affects, developed with your team.
You'd leave with:
- A short set of principles the team agrees on
- A quick check before a new feature ships
- An owner and a measure for each
Social impact review
Your product or one feature, before it ships: who it lands on, and what to change.
You'd leave with:
- The blind spots the team is too close to see
- Changes to make before launch, each with an owner and a measure
- The financial and human return, in your buyers' terms
Social impact and business advisory
An ongoing partner for the human side, as you build.
You'd leave with:
- Someone in the room for decisions that land on people
- Help making the case to buyers and investors
- A closer look at high-stakes features before they ship
When to call me
While you're building.
The earlier we start, the easier it is to build in. Already in market? We start with the feature that carries the most risk, or before your next big launch.
"I've studied what business and technology have gotten wrong, again and again."
I spent two decades inside Fortune 500 and Fortune 1000 companies, building the first social impact and ESG strategies at Kohl's, Victoria's Secret and Advantage Solutions. I know what your enterprise buyers will ask, because I used to be the one asking.
Your team knows how to build it. I help you see who it lands on, and build that in from the start.
Fair questions
Are we too early?
Early is the easiest time to build it in. Principles set now shape every feature after.
Will this slow us down?
The work fits your build. We start with the feature or decision that carries the most risk.
What do you look at that we might miss?
How to prevent unintended harmful impacts on people, communities and society. Most come from untested assumptions: every product is built on assumptions about the people it touches, and most never get tested with them. I look at who your product makes calls about, whether those people helped shape it, and what happens when someone you didn't design for uses it.
Building AI that should help people?
Let's make sure it does, while it's still easy to build in.
Using AI at your company? See how I work with companies