AI Product Manager · Innovation Leader · Emerging Technology
Will Murphy is a Silicon Valley innovation and product leader who has spent 15+ years shipping new products and 10+ years shipping new AI products.
Background
I'm an AI Product Manager, putting AI to work since 2016. Currently at ServiceNow. I've led 0-to-1 AI product bets at Intuit's Futures Group and Data.ai Labs (formerly App Annie).
Earlier, I was founding Head of Product at Talla, an enterprise AI chatbot startup that was acquired, and founded MessagePath, a top-5 AI writing assistant on G2. As a founding member of FedEx Innovation Labs, the company's former zero-to-one group, I co-founded SenseAware, its pioneering IoT analytics venture, led a partnership with eBay to launch a new service, and originated the concept behind FedEx DataWorks.
I coined the "probabilistic product" in 2016, am an OnDeck Fellow, hold multiple AI and IoT patents, and have spoken on AI and innovation at Stanford, MIT VLAB, and the Global AI Conference.
Current Work
Social bookmarking and search reimagined. Save, organize, and discover ideas across the web with a community-driven approach to curation and knowledge sharing.
Visit IdeaFeedA deep dive into the unique design, philosophies, and product management challenges of building AI-driven software.
Learn more
A fully self-contained chess engine running inside a single HTML page. Play against an AI opponent. Built with vibe coding.
Play now
A platform to share, discover, and run GenAI prompts. Making prompt engineering accessible and collaborative for everyone working with large language models.
Visit PromptHubFuture Explorations
An interactive simulation sandbox where small business owners and enterprise leaders can "play out" scenarios like product launches, pricing changes, or market shifts in a generative environment. Moves beyond static dashboards into dynamic, explorable world models for business decision-making.
A new product design paradigm for software built to serve AI agents as the primary user. When the "customer" is an autonomous agent rather than a human, the conventions of UX, onboarding, pricing, and interfaces break down entirely. Exploring what it means to design products where agents discover, evaluate, purchase, and operate software on behalf of humans.
An adaptive interface layer for LLM outputs that morphs based on the intent of the response and the user's cognitive context. Instead of uniform chat bubbles, the UI automatically generates the right interaction pattern, a decision tree, a comparison table, a guided flow, based on what the AI is helping the user accomplish.
A research tool that goes beyond fact synthesis to specifically surface counter-arguments, anomalies, and dissenting evidence in any body of knowledge. Designed to prevent groupthink by mapping opposing viewpoints with full provenance, helping researchers move from "answers" to genuine understanding.
A tool that visualizes why an AI agent failed during complex, multi-step web tasks and allows a human to "steer" it back on track. Focuses on making agentic workflows more reliable in messy, real-world environments through human-in-the-loop debugging and probabilistic design principles.
A focus-management agent that uses live signals from browser activity and wearable sensor data (heart rate, movement patterns) to proactively suggest breaks, deep work modes, or task switching. Combines edge AI processing with augmented intelligence principles to optimize cognitive performance throughout the workday.
Knowledge Sharing
Talks & Media
Previous Speaking Engagements