The team

In 2002, the Sloan Consortium (Sloan-C) brought Keith on as Marketing Director and later promoted him to Chief Operations Officer. In those roles, he helped transform the organization from a collection of scattered research projects into the powerhouse in online education it is today. His leadership drove marketing and technical innovation, as well as data-informed product growth, helping Sloan-C expand from 180 member institutions to more than 1,800—representing approximately 80% of the market—along with more than 200 corporate training divisions. During this period, annual revenues grew from $0 to $1.5 million, while grant funding increased from $20,000 to $12 million.

Now known as the Online Learning Consortium, Sloan-C was a nonprofit start-up positioned at the center of the technological revolution transforming education in the early 2000s. The organization explored new ways to extend education to previously hard-to-reach communities, embracing emerging technologies ranging from mobile platforms to innovative software and online learning systems.

He also played a key role in the Sloan Semester, a landmark 2005 emergency initiative that used online education to replace courses for tens of thousands of students displaced by Hurricanes Katrina and Rita. The effort demonstrated the potential of online learning not only as an emerging technology, but as a powerful and immediate solution to large-scale disruptions in traditional education.

In addition to serving as the COO of Sloan-C, Keith worked with experts in the field on various projects including research on blended learning, business strategies in online education, marketing online programs, and emergency asynchronous learning networks (ALN).

He has published in education-technology journals: “Online Student Profiles: Stay-at-Home Mothers and Work-at-Home Mothers” in the eMentor Journal of Educational Technology, and “Online Program Marketing Budgets” in Distance Education Report.

Between 2020 and 2023 he spent three years building an education-focused generative-AI platform — a round-the-clock tutor with guardrails against inaccuracy, voice-based mock interviews with detailed feedback, and content generation that cut course-creation time from weeks to minutes.

Forward Deployed AI Engineer at Tribe AI, where the work is helping enterprise teams move from isolated LLM features to measurable, goal-driven agent systems. Before that: a CoALA-based agent memory system at Magnifi by TIFIN (500K+ users, $4B linked assets), and a modular AI platform at Johnson & Johnson used by tens of thousands of data scientists. Ten years shipping production AI systems; before that, a COO who scaled an organization 10×. The architecture lens and the operating-model lens are the same lens.

"I shipped a multi-agent platform last year. I typed about 5% of the code." — "I architected it. Agents built it."

Author of Unlocking Data with Generative AI and RAG (Packt, 2 editions) — the second edition covering graph-powered RAG agents, memory pipelines, semantic caching, and agentic system design. Host of The Memriq AI Inference Brief: weekly, two editions, built for practitioners and the executives who fund them.

The book has two published editions from Packt: the first in September 2024 at 346 pages, and the second in January 2026 at 606 pages, expanded into agent memory systems, graph-based retrieval and agentic system design. Its foreword is by Shahul Es, founder of the open-source evaluation library Ragas. He is currently writing the third edition.

Memriq is not being built to replace human purpose — it is being built to protect it: agents take on the administrative weight so people get their time, attention, and energy back for what actually matters. Alloy is that thesis pointed at one thing. Preparing for a certification is mostly overhead — working out what to study, building the practice, tracking what has decayed — and none of it is the part only you can do. The agents carry that so your hours go on the material and the exam. The CAO role here is deliberately minimal — set direction, review exceptions, approve what needs approving. The agents have the reins. The system is the work.