Resume

David

AI Systems Builder · Agents · Automation · Generative Media
Australia (AEST) · David_plos@hotmail.com· github.com/Davidplos· LinkedIn

Download PDF Contact

PROFILE

Hands-on builder of practical AI systems: multi-agent environments, self-hosted automation, generative media pipelines and AI-assisted applications. I design, integrate and operate end-to-end systems — several running daily, one producing monetised revenue. Comfortable across the full stack: LLM orchestration, workflow automation, local inference, media generation and deployment.

CORE CAPABILITIES

  • ▸Multi-agent AI orchestration & autonomous task execution
  • ▸LLM application development (cloud & local inference)
  • ▸Workflow automation with n8n, webhooks & API integration
  • ▸MCP (Model Context Protocol) tool integrations
  • ▸Browser automation & web scraping
  • ▸Generative media: image, video, voice, music
  • ▸Programmatic video production (Remotion/FFmpeg)
  • ▸AI-assisted software development & agent coding tools
  • ▸Python scripting, dashboards & data pipelines
  • ▸Self-hosted infrastructure (Linux VPS, Docker)

PROJECT EXPERIENCE

Hermes Multi-Agent AI Environment

Ongoing

Personal multi-agent environment: orchestrator + specialist agents, 152+ skills across 16 categories, MCP tool access, browser automation, research and code generation.

Python · MCP · Playwright · LLMs

AI Content Production System — YouTube

Ongoing

End-to-end faceless content pipeline operating 4 channels (1 monetised, $200–550/mo): research → script → AI voice → generated visuals → programmatic edit → publish.

ComfyUI · TTS · FFmpeg · Hyperframes

Local Generative AI Lab

Ongoing

Local-first generation on RTX 5070 Ti: ComfyUI image/video workflows, llama.cpp local LLMs, multilingual voice synthesis (PT/DE/FR/JA/NL/EN-AU), music generation.

ComfyUI · llama.cpp · GGUF · FFmpeg

Self-Hosted Automation Infrastructure

Ongoing

VPS running n8n workflow engine, Baserow database, Dograh voice platform; webhook/API integrations connecting LLMs to business processes.

n8n · Baserow · Docker · Linux

AI Voice Receptionist

In development

Voice agent prototype: caller audio → STT → LLM agent → n8n tool actions (calendar/CRM) → TTS response. End-to-end flow demonstrated; not production-hardened.

Dograh · n8n · STT/TTS · Webhooks

Programmatic Video Pipeline

Completed

Code-driven video compositions with Remotion/Hyperframes/OpenMontage + FFmpeg rendering; version-controlled, reproducible output from structured data.

Remotion · TypeScript · FFmpeg

TECHNOLOGY

AI / Agents
Claude, LLMs, agentic frameworks, MCP, multi-agent orchestration
Automation
n8n, APIs, webhooks, browser automation
Development
Python, JavaScript/TypeScript, Git, VS Code, Cline, Claude Code, Codex
Generative media
ComfyUI, image/video/voice/music generation, FFmpeg, Remotion
Infrastructure
Linux, VDS, Docker (where used), self-hosted services

CERTIFICATIONS

None listed — my evidence is working systems, not certificates. Happy to complete role-relevant certifications on request.

CONTENT / MEDIA EXPERIENCE

  • ▸Operate 4 YouTube channels through the AI content production system
  • ▸1 monetised channel generating $200–550/month demonstrated revenue
  • ▸Multilingual voiceover production: PT · DE · FR · JA · NL · EN-AU
  • ▸Full-stack single-operator workflow: research, scripting, voice, visuals, edit, thumbnail, metadata, publishing

CURRENTLY LEARNING

  • ▸Hardening the AI voice receptionist for real inbound call volumes
  • ▸Deeper ComfyUI workflow engineering (control nets, advanced sampling chains)
  • ▸Production observability for agent fleets (tracing, evals, cost control)

CONTACT

Open to remote roles, contract projects and collaboration.

Email meSee the work