Self-taught systems builder · Building infrastructure for the agent economy
I started coding at 15 because I've always had a knack for spotting problems and immediately seeing how they could connect to a practical solution. Once I realized tech and coding were the perfect way to put that trait to work, I dove in headfirst. Since then, I've been building solutions to pretty much any problem I actually care about — whether it's something small in my daily life or something bigger on the frontier of how systems operate.
Over the years that curiosity has taken me through just about every corner of tech: blockchain and crypto-economics, AI, Python scripting, bash, Linux environments, frontend work, web design, distributed systems, multi-agent coordination, AI infrastructure, economic systems, protocol design, safety, and governance. If I have to do something more than twice, I automate it. That habit alone has saved me countless hours and taught me a ton about building reliable, repeatable systems.
Everything I know about programming and systems is self-taught through building. I never followed a formal path — I just kept shipping things that solved real problems for me or that I thought would matter. That hands-on approach gave me a broad, connected view across the stack, from low-level runtime enforcement to high-level economic coordination. It also sharpened my ability to move fast: spot an issue, prototype a fix, and iterate until it works.
| Language | Level |
|---|---|
| Python | ██████████ Primary |
| TypeScript | ████████░░ Strong |
| JavaScript | ████████░░ Strong |
| Go | ██████░░░░ Working |
| SQL | ██████░░░░ Working |
| Bash | █████░░░░░ Scripting |
| Rust | ███░░░░░░░ Learning |
Distributed Systems · Multi-Agent Systems · AI Infrastructure · Economic Systems · Protocol Design · Safety & Governance
The agent space is exciting but still fragmented — different frameworks, different protocols, different ways of talking to tools and each other. I'm building the infrastructure that will let autonomous AI agents actually work together in the real world.
My most important project right now. A layer that lets any AI agent connect to any tool, any API, or any other agent directly from your terminal (or anywhere else). It sits in the middle as an adaptive semantic interoperability layer:
- Translates between protocols — A2A, MCP, ACP, and more
- Self-correcting mappings — keeps integrations healthy over time
- Agent discovery & coordination — gives agents a clean way to find each other and work together
It's the kind of practical bridge the whole ecosystem needs right now.
| Repository | Description |
|---|---|
| monorepo_core | Full execution environment with SDKs, enforcement, economics, and monitoring |
| guard_backbone | Deterministic runtime firewall — checks every action before it runs |
| translator_middleware | Protocol translator enabling cross-framework agent collaboration |
| Repository | Description |
|---|---|
| a2a_coordination | Negotiation and coordination layer for agents |
| task_formation | Dynamic team formation and collective task handling |
| a2a | Economic coordination engine for agent ecosystems |
| metrics | Causal impact and synergy evaluation |
| influence_system | Trust-aware influence modeling |
| humanlike_agents | Persistent identity frameworks |
| environmental_awareness | Context awareness for agents |
| Repository | Description |
|---|---|
| enforcement_layer | Multi-layer guardrails for safe autonomous systems |
| identity_system | Persistent identity and authority for agents |
| actionable_logic | Executable governance and policy frameworks |
| economic_autonomy | Resource and budget management for agents |
| economic_layer | Governance layer for agent ecosystems |
| simulation_layer | Governance testing environments |
| self_improvement_governance | Controlled self-improving agents |
| scorring_module | Decision evaluation and risk scoring |
I keep a few real-world agents running to keep the theory grounded:
| Repository | Description |
|---|---|
| email_agent | Cold outreach automation |
| leads_agent | Global lead discovery |
| contentagent | Personalized content generation |
| mega2 | Lumina Utility Bill Optimizer |
| landing_page | Product landing page — live |
| cool-LOC | LOC counter for swarms |
| cloud_demo | Cloud deployment demo |
| Strength | Description |
|---|---|
| Problem-to-solution wiring | I see connections others miss and turn them into working code quickly |
| Broad systems thinking | Comfortable across the entire stack — languages, tooling, and domains |
| Self-taught execution speed | I ship, learn, and refactor in the same motion. No waiting for permission or perfect docs |
| Automation-first mindset | Everything repetitive gets turned into reliable infrastructure |
| Vision with follow-through | Not just theorizing about the agent economy — building the actual layers it needs to scale |
Autonomous agents are going to become a major layer of computing. For that to happen safely and at scale, we need more than clever models. We need:
- Deterministic safety — not probabilistic guardrails
- Real coordination — across frameworks and protocols
- Interoperable tools — a connected ecosystem, not walled gardens
- Governance that evolves — systems that can adapt their own rules responsibly
That's the substrate I'm focused on creating. Not hype, just solid infrastructure that lets agents discover each other, talk across protocols, share tasks, and operate inside shared systems without breaking things.
Self-taught. Systems-driven. Building the connected agent future.