Our Noble Purpose
We help leaders navigate today and lean into tomorrow, helping them access more intelligence, from more sources, in service of creating more value, all accelerated through the adoption and use of AI.
At Hi™, we help leaders and professionals build a productive AI mindset and a practical toolkit to get the most from their Intelligence Ecosystem™. Our approach is built on a modular and customizable learning journey, tailored to each client’s unique needs and strategic imperatives. Through discovery-based learning and immediate practical application, we ensure that AI adoption becomes a catalyst for human flourishing, not a threat to it.
Our Team
Richard Hodge
Co-Founder
Richard has spent more than five decades building companies that change how leaders learn, how sales forces sell, and how organizations adopt new technology.
His early career ran through enterprise software and energy, with stops at AT&T, Occidental Petroleum, and Execucom, where he helped shape the first wave of decision support systems. He was part of the Novell turnaround team, an experience that hardened his conviction that curiosity and experimentation beat mandated training every time.
In 1990 he co-founded Advantage Performance Group, and in 1994 he founded The Real Learning Company. Over the following dozen years, RLC built and delivered discovery-based sales and leadership programs for more than 300 Fortune 1000 companies, working alongside many of the field’s best-known authors and researchers.
In 2006, both companies were acquired by BTS Group, where Richard served as Executive Vice President and Global Partner for 15 years. He helped steer the firm’s global learning practice through the shift to virtual delivery and co-developed The Collective Intelligence Simulation, built on Harvard researcher Linda Hill’s work on leading innovation.
Today Richard is Chairman of 1st90, a Senior Vice President at Advantage Performance Group, and co-founder of Harness Intelligence. He is the architect of the 3 Gs of Responsible AI Use (Guidance, Guardrails, and Governance), a practical framework that gives teams room to experiment with AI while keeping the organization safe.
Notable Work
Co-founded: Advantage Performance Group (1990), The Real Learning Company (1994)
Book: The Mind of the Customer (McGraw-Hill, 2006), with Lou Schachter
Simulations: The Collective Intelligence Simulation, built on Harvard researcher Linda Hill’s work
Framework: The 3 Gs of Responsible AI Use (Guidance, Guardrails, Governance)
Recognition: Multiple industry honors for program excellence
Education: BBA Finance and MBA, University of Southern California
Jim Perry
Co-Founder
Jim is a learning designer, facilitator, and builder. Over 25 years he has turned leadership and psychological research into hands-on simulations and programs that have reached tens of thousands of professionals across the Fortune 100.
He spent 17 years at Kinko’s, rising from Operations Manager to Director of Training. He built the company’s technical training curriculum for 20,000 field employees and managed a 300-person field training network through a decade of rapid national expansion. Afterward he ran Business Video Express, a mobile multimedia training company, where he first met and began collaborating with Richard.
He then spent 17 years at BTS Group, rising to Vice President and Head of Leadership Curricula. He led the team of instructional designers behind flagship simulations and learning maps used by more than half the Fortune 100.
He is best known as the lead designer of the Multipliers simulation, developed with Liz Wiseman and based on her New York Times bestseller of the same name. The simulation puts leaders in a controlled environment where they can feel, in real time, the difference between diminishing and multiplying their team’s intelligence. Research done alongside the rollout found that leaders operating as multipliers access roughly twice the intelligence of their teams compared with leaders who diminish them, and tens of thousands of managers have now worked through it.
At Hi, Jim co-designed the Intelligence Ecosystem framework and the IntelliPRINT assessment system, developed the PGA prompting framework (Persona, Goal, Ask), and builds most of the technology behind the company’s products. He is a long-time Advantage Performance Group thought leader partner and a master facilitator for The Multipliers Simulation, Symphony, The Collective Intelligence Simulation, and many other flagship leadership programs.
Notable Work
Authored: The Multipliers simulation, with Liz Wiseman (based on her New York Times bestseller)
Developed: The PGA prompting framework (Persona, Goal, Ask)
Co-designed: Intelligence Ecosystem framework, IntelliPRINT, AI Readiness Pulse Check
Master facilitator: Multipliers, Symphony, The Collective Intelligence Simulation, and other flagship programs
Recognition: Multiple industry honors for program excellence
Education: B.A. Philosophy, Doane University
Our Trusted Advisors
Erin Perry
DAYTA Marketing
Jon Hodge
APG
Kelvin Yao
APG
Kristina DiStasio
Sr. Partner APG
Mary Steiner
Sr. Partner APG
Peg Ruppert
Sr. Partner APG
Praful Saklani
Pramata
Tim Blakesly
USC Marshall Executive Education
What We’ve Built
We don’t just teach AI. We build it. Behind our programs sits a working portfolio of software our team has designed, built, and runs in production, from autonomous agents to full-stack web apps. A few examples:
The Guide
A web app that runs our live AI workshops.
The Guide keeps a continuously updated profile of every major AI platform, including ChatGPT, Copilot, Gemini and Claude, tracking the exact menus, buttons and steps each one uses. When a participant opens their page, it assembles instructions for the specific tool in front of them and folds in the client’s own examples and content. Everyone in the room follows accurate, personalized walkthroughs, even though every tool works a little differently.
Python · PHP · Flask
Courier Open Source
An email and calendar client built for AI agents to operate, not people.
Most tools that connect an AI to email do it through an API or an MCP server that mirrors the buttons of a human email app, which forces the AI to work the way a person clicks through a screen. (MCP, the Model Context Protocol, is the emerging standard for how AI assistants plug into outside tools.) Courier takes a different path: it speaks directly to the open standards behind email and calendars, JMAP and CalDAV, and presents them as actions an agent actually reasons about, such as “what has arrived since I last checked.” In our own automation that did the same work in roughly half the steps and half the cost. We are releasing it publicly under an MIT license so other builders can use it.
Python · JMAP · CalDAV · MCP
Pulse Check
A web app that measures how ready an organization is to adopt AI.
Behind the scenes it routes each question to a different size of AI model depending on how hard the task is, using a fast, inexpensive model for simple classification and a more capable one only when the reasoning calls for it. That tiering keeps the analysis sharp while holding the running cost down, the same discipline that makes AI products affordable at scale.
Next.js · PostgreSQL · Claude API
Meeting Agent
An AI participant that joins a Zoom call and takes part by voice.
It chains several systems together in real time: speech recognition to hear the room, a language model to decide what to say, and speech synthesis to say it back, all bridged into the live meeting. Real-time voice pipelines like this are genuinely hard to build, and most AI demos stop well short of them.
Python · Pipecat · Deepgram · Cartesia
Hi Team Hub
Our own internal operating system, built as a private web app.
It pulls the many separate tools a business runs on, including email, calendar, documents and project notes, into a single dashboard, and it runs on hardware we control rather than a third-party cloud. It keeps the whole company coordinated from one screen, and it shows we can build the same kind of unified workspace for a client.
Next.js · TypeScript · SQLite
The Reinvented T3
A new take on train-the-trainer, where learning the program and learning AI happen together.
Each candidate facilitator works hands-on with their own company’s AI tool, whether that is Copilot, ChatGPT or another, to learn the program from the inside out. Starting from raw session transcripts and source materials, they build their own personalized facilitator guide rather than memorizing someone else’s script. The result is two kinds of mastery at once: deep command of the program, and real fluency with AI, since along the way candidates create their own agents, documents and job aids. Throughout, the same Claude that helped build and now runs our programs works alongside each candidate as a personal learning coach.
Claude · Copilot · ChatGPT
HiBots
A suite of custom AI assistants we built before consumer AI offered them.
Long before ChatGPT let anyone create a custom GPT, we were building purpose-built assistants, each given its own knowledge, personality and job, and putting them in front of real users. HiBots was our early proof that a general model could be shaped into a dependable specialist that does one thing well. The lessons from it run through everything we have built since.
Python · Streamlit · OpenAI
Joan
A long-running autonomous agent that triages email on its own, around the clock.
Joan wakes every fifteen minutes on a server we run, reads what is new, and decides what to do with each message. It handles the routine, files what it can, and surfaces only what truly needs a person, logging every action so nothing happens in the dark. We built it deliberately careful: it never deletes mail and every move can be undone. Joan was our experiment in trusting an agent to work unattended for weeks at a time, and it does. (It runs on Courier, the email client above.)
Python · Claude · Courier · launchd
IntelliPRINT
A new assessment that profiles how individuals and teams work with intelligence.
IntelliPRINT measures how a person or a group gathers, shares and acts on intelligence, then turns that into a practical profile they can put to work. It has grown through several generations, and the current version is being rebuilt on a modern web platform designed to scale from a single individual to a whole organization. It is the assessment engine behind our work on the Intelligence Ecosystem.
Next.js · Vercel · OpenAI
Across the portfolio we work in Python, TypeScript and PHP, with modern web frameworks, SQL databases, and AI models from Anthropic and OpenAI connected through the Model Context Protocol. We handle the whole lifecycle ourselves: design, build, hosting and day-to-day operation.
Let’s explore what’s possible
Ready to turn AI uncertainty into strategic advantage? We’d love to hear about your organization’s goals.