About
I am the Co-Founder and CTO of EYwALink, where we specialize in turning business concepts into concrete value. By bridging theoretical models and practical, production-ready AI agentic systems, I help organizations operationalize Agentic AI at scale. My work combines hands-on leadership in agentic engineering with strategic enterprise architecture consulting.
Topics: Enterprise Architecture · Agentic AI Strategy · Digital + AI Transformation · Software Engineering
Posts
From a Security Incident to Reconsidering Enterprise Agent Boundaries
Why agent boundaries are the new enterprise security perimeter, the real cross-tenant incident that exposed the gap, and the isolation tiers that actually hold. The application boundary is only as strong as the identity underneath it.
The Cost of Running an AI Lab in Local AI
Running Eywalink, an AI company, is cheap at the model layer and expensive everywhere else. Over eight months, one Mac and a server of RTX...
…From One Agent to a Company of Agents
A coding agent is a tool you invoke. A company of agents is a system that runs. After a year of running EyWALink and its multiple client tenants as agent-run companies — every role an agent, every task an issue, every invocation a heartbeat — the lesson is that the model was never the hard part. The control plane was.
…Effective diagram tools for agentic enterprise architecture
AI agents generate enterprise architecture content at a rate no diagram toolchain can keep up with. In Agentic Executable EA, agents produce capability maps, application...
…When Your Agent Says "Done"
TL;DR: My own verification script marked a half-finished project complete. That ended the trust-based workflow. Now completion is an observable fact: a task queue, executable...
…Open Agile Architecture Overlay is released for agile digital transformation
TL;DR: Two months ago, I published Executable Enterprise Architecture, Agentic-TOGAF-ADM, where I showed how the ArcKit open-source project turned the TOGAF ADM into agent-readable slash...
…How I Built a Digital Twin of Myself That Actually Scales
Four iterations of building a digital twin of myself, an AI mirror of who I am, what I know, and how I think. It compounds over time without exploding context windows or leaking data to the cloud.
…Why Zero Lock-In Matters for Enterprise AI
TL;DR: Every CTO has heard the pitch. “Just use our managed AI platform; it’s the easiest way to get started.” They’re right about the easy...
…Private AI Must be part of the Enterprise AI Strategy
TL;DR: The real strategic question isn’t “Which LLM vendor?” It’s “Who owns your AI intelligence?” Exclusive reliance on public cloud AI introduces compliance exposure, fragments...
…Managing Context Overflow in AI Coding Agents
Context overflow is an architecture problem. When coding in a large project codebase, context windows kill agent sessions mid-task. Here's how I architect handoff files, subagent delegation, and task decomposition to keep coding agents productive across session boundaries.
…Bridging the E2E Testing Gap: Introducing SuperWeb Testing
TL;DR: End-to-end testing is supposed to be the last line of defence before users hit broken features. In practice, it is often the first thing...
…Observability Stack: Architecture for multi-agent AI systems
TL;DR: Autonomous AI agents are learning to write code, debug themselves, and orchestrate dozens of parallel workstreams. They’re fast. They’re tireless. And they’re producing systems...
…Manage Your AI Agents at Enterprise Scale
TL;DR: Enterprises are deploying hundreds of AI agents across their operations, but most lack a systematic way to govern, secure, and orchestrate them. Recently I...
…Executable Enterprise Architecture - Agentic-TOGAF-ADM
TL;DR: Are you an executive leader without an architecture background, frustrated by the disconnect between your business goals, operational realities, and technology landscape? You know...
…Practice of AI agent workflow patterns
TL;DR: Building software with AI agents involves an iterative, feedback-driven looping process, instead of writing prompts manually. In my open-source project loop_engineering_factory, I’ve been developing...
…Why I build my local AI with less spend on AI Subscriptions
TL;DR: 2026 is shaping up to be the year of Agentic AI. As open-source LLMs mature for local deployment, the line between model and harness...
…Buy Over Build? Rethinking IT Strategy in the AI wave
TL;DR: For two decades, “buy over build” was enterprise IT’s default constitution. It was rational: dev talent was scarce, legacy maintenance was brutal, and vendors...
…The Impossible Triangle: Speed, Memory and Intelligence
TL;DR: As AI shifts from rule-based logic to pattern-driven LLMs, knowledge is no longer stored as explicit rules; it’s compressed into statistical intuition across billions...
…Private AI Is Production-Ready
After billions of tokens through local silicon, private AI on open source stacks is ready for production. Here's the infrastructure, the tooling, and the workflows that work.
…