Organizational compression: AI as a new operating leverage for the enterprise
Organizational compression shows why AI becomes operating leverage for more agile enterprises that decide and execute with less friction.
AI strategy / solution architecture / delivery
Strategy, architecture and reliable AI systems for enterprise teams.
I help enterprise teams and organizations define AI strategy, solution architecture and reliable agentic workflows for real, complex and regulated environments.
Current focus
- BuildingPersonal AI assistant (AIA), an agentic orchestration platform built on Hermess to automate my own workflowsactive
- Last shippedSpecDraft RunnerLocal Streamlit dashboard for exploring speculative decoding with an Ollama-served target model and DSpark draft-head assets.live
Expertise
A pragmatic bridge between strategy, architecture and delivery.
01
Generative AI strategy
From opportunity discovery to pragmatic roadmaps, governance and measurable enterprise adoption.
02
Solution architecture
Secure, scalable architectures connecting cloud platforms, data, AI models, integrations and operating constraints.
03
Agentic workflows
AI agents, automation patterns and human-in-the-loop systems designed for real business processes.
04
Digital health & SaMD
Technology strategy for regulated software contexts where safety, traceability and evidence matter.
Articles
Analysis, notes and applied perspectives on enterprise AI.
Hermes Agent optimization is a systems architecture discipline
Hermes Agent becomes expensive or unreliable when context, tools, skills, memory, subagents and scheduled jobs are left unmanaged. The right optimization model is architectural: route work, limit context, govern background automation and reserve expensive reasoning for the tasks that need it.
Chinese open models are forcing a new economic logic for enterprise AI
The real disruption from Chinese AI models is not just capability. It is the combination of lower cost, greater deployability and stronger provider optionality that can reshape enterprise AI architecture.
Case Studies
Applied case studies, documented like working projects.
Stanford AI Index 2026
A structured reading companion to Stanford HAI’s 2026 AI Index with a practical overview, chapter drill-downs and compact data tables.
AI Providers & Model Catalogs
A practical catalog of the companies shaping the AI landscape from frontier labs to cloud giants, with provider pages, model lines and quick positioning notes.
The Portal Design System
How this site is designed and themed: design tokens, a four-theme system including high-contrast and color-safe accessibility modes, typography and a shared component language.
Pi Coding Agent
An in-depth look at Pi Coding Agent: the open-source, fully customizable agentic coding tool that pushes beyond Claude Code with multi-team orchestration, peer-to-peer agent communication, and extensible agent harness.
Contact
Let's discuss the next AI system you want to make reliable.
If you are evaluating an AI strategy, an application architecture or an agentic workflow to bring into production, we can talk. I work best where technical judgment, pragmatism and real-world constraints matter.
Complexity is anything related to the structure of a software system that makes it hard to understand and modify the system.