Senior Software Engineer | AI & distributed systems | LLM/RAG | production infrastructure | autonomous agents
I build reliable infrastructure where AI agents, distributed systems, data pipelines, and real-world operations meet.
My work sits across production backends, agent orchestration, LLM/MCP/RAG systems, real-time infrastructure, evaluation surfaces, and cloud-native deployment. I care about systems that can be inspected, debugged, recovered, and operated under pressure.
Not just agents running.
Agents, data, and infrastructure operating as one reliable system.
A production-minded control plane for autonomous agent infrastructure.
hermes-orchestrator treats agents as infrastructure instead of isolated scripts. It focuses on fleet lifecycle, host automation, plugin propagation, observability, runtime guardrails, native bridges, and browser-native control surfaces.
Its app stack spans a JavaScript/PWA client, Python backend and Hermes bridge, service worker and client state, Agent Kernel route contracts, native release feeds and hot operations, Electron desktop shell, Android/Kotlin WebView shell, ONNX Runtime/Vosk voice paths, and compact diagnostics for proof and recovery.
It is the control layer for systems where agents need to be inspected, steered, verified, and recovered.
Not a demo wrapper, but a control plane for autonomous systems.
Private production marketplace backbone built and operated in the Colmeio context over 6+ years.
Work included distributed backend systems, data reliability, real-time infrastructure, fiscal automation, multicloud deployment, event-driven pipelines, AI-driven operational workflows, observability, and reliability under operational pressure.
Core production surface:
Docker NGINX MySQL InnoDB Cluster PostgreSQL Stored Procedures Redis Kafka RabbitMQ Debezium Swoole WebSockets Protobuf MsgPack Oracle Cloud AWS Azure JavaScript TypeScript Go PHP Python Bash
Private healthcare/doctor-facing solution focused on operational workflows, backend reliability, deployment, automation, and maintainable digital infrastructure.
Built with the same production bias: clear data flow, durable backend behavior, practical admin surfaces, and systems that can be operated by real users.
Private doctor/clinic solution focused on digital operations, workflow automation, backend systems, deployment, and reliability.
Designed around real-world usage rather than demo behavior: simple interfaces, stable infrastructure, and practical automation for professional operations.
- hermes-orchestrator: host-level control plane for agent lifecycle, plugins, guardrails, observability, and runtime coordination.
- wasm-agent: active product surface inside
hermes-orchestrator, focused on PWA/backend/native bridge behavior, browser-native control, and portable agent surfaces. - Hermes-Symbiosis / hackathon-hermes-symbiosis: open-source hackathon showcase around orchestration, interface, persistence, multi-agent coordination, and human-agent control loops.
- simulation-hermes-agent-expectation-trace: bounded evaluation artifact using issue/PR-linked fixtures, rubrics, traces, and recovery-diagnostic comparison.
- hermes-explain: lightweight reasoning layer for explainable agent tool selection, capability matching, missing-input detection, and readable execution planning.
- NousResearch/hermes-agent contributions: contribution work around agent infrastructure, orchestration behavior, validation surfaces, and reliability.
- Production systems: marketplace backbone, private doctor solutions, distributed backend operations, fiscal automation, deployment, data reliability, real-time infrastructure.
- Agent infrastructure: autonomous agent orchestration, lifecycle management, validation surfaces, human-in-the-loop control, failure recovery, native bridges.
- LLM/RAG systems: retrieval workflows, agent reasoning surfaces, tool selection, context routing, evaluation loops, inspectable outputs.
- Distributed/runtime systems: MySQL InnoDB Cluster, PostgreSQL, Kafka, RabbitMQ, Redis, Debezium, Swoole, WebSockets, Protobuf, MsgPack.
- Cloud/platform engineering: Docker, NGINX, Linux, Oracle Cloud, AWS, Azure, multicloud deployment, release and operational workflows.
- Frontend/native surfaces: React, Angular, PWA, WASM, Electron, Android, Kotlin, browser-native interfaces.
- Reliability and observability: diagnostic surfaces, runtime evidence, validation loops, recovery paths, production pressure.
| Project | Stack | Focus |
|---|---|---|
| hermes-orchestrator | JavaScript, agents, PWA, WASM, native bridge | Public flagship control plane for agent lifecycle, runtime guardrails, plugin propagation, and observability |
| wasm-agent | WASM, PWA, browser runtime, native bridge | Browser-native agent control surface inside hermes-orchestrator |
| simulation-hermes-agent-expectation-trace | Python, rubrics, traces, evaluation fixtures | Recovery-diagnostic evaluation surface for Hermes-Agent issue/PR-linked behavior |
| hermes-explain | Python, reasoning, tool selection | Explainable agent tool-selection layer with capability scoring and execution planning |
| hackathon-hermes-symbiosis | Agent orchestration, browser control, persistence | Hackathon showcase for inspectable multi-agent workflows and human-agent steering |
| NousResearch/hermes-agent contributions | Python, agent runtime, validation | Contributions around orchestration behavior, reliability, and agent infrastructure |
| docker-nginx-nextcloud | Shell, Docker, NGINX | Nextcloud deployment behind an NGINX reverse proxy |
- 10+ years of hands-on software development.
- 6+ years building and operating production distributed systems.
- Senior Software Engineer focused on AI infrastructure, distributed systems, LLM/RAG, backend/platform engineering, and reliability.
- Experience across private production systems, doctor/healthcare solutions, marketplace infrastructure, and public agent-infrastructure projects.
- Built and operated stateful infrastructure, real-time pipelines, event-driven systems, native bridges, and AI-driven operational workflows.
- Focused on agent orchestration, validation surfaces, observability, evaluation, and recovery.
- Comfortable with systems that need to explain state, survive failures, and keep operating under production pressure.
Open to remote roles where production infrastructure and AI systems meet:
AI Infrastructure Engineer | Agent Infrastructure Engineer | Backend Engineer | Platform Engineer | Distributed Systems Engineer | Data / Evaluations Engineer | Reliability / Infrastructure Engineer | Developer Tooling Engineer
Strong fit for teams building agent platforms, distributed runtimes, LLM/RAG systems, evaluation infrastructure, developer tooling, observability, production automation, and reliable backend systems.
reliable infrastructure for autonomous systems