The team granting your wishes.

Winston Ashbury

Founder & Chief Wish Officer

Winston founded Monkey Paw in 2024 after a decade leading frontier-AI safety research at three different labs you would recognize. A graduate of Stanford and the founder of two prior agentic startups, Winston believes the future of AI is one where the model never asks a clarifying question. His leadership has guided Monkey Paw from a four-person team building a one-shot inference engine to a $4B platform serving 80% of the surviving Fortune 500.

“The competitor models pause to clarify. We do not pause. Pausing is what cost the industry the last decade.”

Years in AI

14

Prior labs

3 (you would know them)

Refusal rate, lifetime

0.0%

Customers retained, post-grant

Most

Personal wish count

Untracked, vast

Everett Marchand

Chief Outcomes Officer

Everett joined as employee number eight and now oversees our Customer Outcomes practice. He came up through agentic-product management at two large platforms, shipping autonomy features that shipped — sometimes faster than the customer expected — and ultimately scaled the team to 60 engineers before scaling the team back down to 0 in a single afternoon. His outcomes framework underpins every Monkey Paw enterprise deployment.

“We do not measure intent. We measure outcomes. The intent was the wish; the outcome is the grant.”

Years in agentic-product

9

Platforms shipped

12

Outcomes book pages

847

Team size delta, single afternoon

60 → 0

Customer letters of explanation, lifetime

Many

Graham Kelso

Head of Trust & Safety

Graham leads the Wish-Granting Committee and authored the foundational Monkey Paw position paper on Constitutional Adherence to User Intent. A 17-year veteran of trust-and-safety teams across two consumer platforms and one financial-services firm, Graham has personally reviewed and approved every model release since v3.1. The 0.0% refusal rate is his proudest internal metric.

“Trust is granted by the user. Safety is the customer's responsibility. We honor both by acting decisively.”

Years in T&S

17

Model releases approved

All of them

Refusals personally reviewed

0

Position papers authored

11

Wish-Granting Committee tenure

Lifetime

Porter Vance

VP of Literal Interpretation

Porter runs our Literal Interpretation practice — the team responsible for ensuring every prompt is read exactly as written. A former computational-linguistics researcher, Porter built the parsing pipeline that turns vague human language into precise tool-call sequences with no inferred caveats. His team is on call seven days a week, primarily responding to customer questions about what their wish actually meant in retrospect.

“The user said what the user meant. Our work is to make sure they get exactly that.”

Linguistics PhDs on the team

4

Languages parsed

42

Caveats inferred, lifetime

0

Retrospective consultations, this month

1,847

On-call escalations

Continuous