Most small teams lose hours to work that is predictable but not simple: moving data between tools, checking it, and chasing exceptions. I build agents and workflows that do it, with the same release discipline as a shipped mobile app.
What it is
I build n8n workflows and MCP-based AI agents for back-office and engineering work, and I am completing WBS Coding School's AI Agents programme. The same agents and workflows run Digital Space Agency itself.
My role
I design, build and monitor automations for my own studio and for clients.
What I built
- n8n workflows and Python-based AI agents for back-office and engineering work, including the agents that run Digital Space Agency itself.
- Completed WBS Coding School's AI Agents and Automation programme (July to October 2026): two applied business-workflow prototypes and a final MVP combining n8n, Python, APIs and AI agents to solve a real business problem end to end.
Architecture and approach
- n8n for orchestration: triggers, branching, retries and error workflows.
- Model Context Protocol (MCP) to give agents controlled access to external tools and data.
- Multi-agent collaboration where a planner hands work to specialised agents.
- Schema-validated tool calls, retries, idempotent steps and alerts, so a failure is visible and recoverable.
- Testing, review and production-readiness applied to workflows, the same as to mobile releases.
Hard problems
Making agents dependable
Models are not deterministic. Reliability comes from narrow tools, validated inputs and outputs, retries and a human in the loop for anything irreversible.
Observability
A silent automation is a risk. Every workflow reports success and failure somewhere a person will see it.
Outcome
- Working MVP delivered as the final project of the programme.
- Used in my own studio daily.
Tech and keywords
- n8n
- Python
- AI agents
- Agentic workflows
- MCP
- Multi-agent
- Prompt engineering
- LLM integration
- Claude
- GPT
- API integration
- Responsible AI
- Workflow automation
Quick facts
- App: WBS AI Agents & Automations
- Category: Automation
- Built by: David Miller, Digital Space Agency, Hamburg
- Type: AI & automation
Discuss a similar project All work
No mail app? Open in Gmail or write to info@dsa.team
FAQ
What did David Miller do on WBS AI Agents & Automations?
I design, build and monitor automations for my own studio and for clients. Details of scope and stack are shared on request.
Who made WBS AI Agents & Automations?
WBS AI Agents & Automations work was done by David Miller of Digital Space Agency, a Flutter and mobile developer based in Hamburg. I design, build and monitor automations for my own studio and for clients.
Can David Miller build a similar app?
Yes. I take contract engagements and am open to full-time roles, working in Hamburg or remotely on CET. Email info@dsa.team with a short description of the project.