12 Agents, One Audiobook
Multi-agent orchestration in practice: from technical book to professional audiobook series
Multi-Agent · FastAPI · React · PostgreSQL · Redis
I had a problem. I knew things — about autonomous AI systems, multi-agent architectures, the mistakes everyone makes building their first agent. But this knowledge was stuck in my head. And as long as it's there, I'm the bottleneck.
The Bottleneck Problem
Knowledge that exists in only one head doesn't scale. Blog posts? No time. Book? Even less. YouTube? I don't like cameras. But one format stuck: audiobooks. A multiplier — it reaches people in moments no other format can.
The Trigger
Antonio Gulli's book on AI agents. Brilliantly written, technically deep, but dense. 400 pages that not everyone can simply work through. Not because readers are dumb — but because not everyone has the background to translate architecture diagrams in their head.
I wanted to make this knowledge accessible. Not water it down, not simplify until it's wrong — but put it in a format everyone understands.
12 Agents, One Production Line
One Prompt
- No quality control
- Average output
- Factually uncertain
- Not optimized for speech
- One-size-fits-all
12 Orchestrated Agents
- Each agent has a clear role
- Quality gates at every step
- Fact-check against source
- Voice style optimization
- Professional production
Research Agent
Analyzes source material, extracts key points.
Story Architect
Plans dramaturgy — how to make MPPT exciting?
Writer Agent
Writes the episode with clear style.
Revision Agent
Checks consistency across all episodes.
Sensitivity Agent
Ensures simplification doesn't become error.
Voice Style Agent
Optimizes for spoken language.
The Multiplier Effect
ElevenLabs with voice cloning. ACX standard: -23 to -18 dB RMS, max -3 dB peak, noise floor under -60 dB. Professional enough for any platform.
But the real point isn't the technology. The point is: I'm no longer the bottleneck. Technical book in, accessible audiobook episodes out. Scalable, reproducible, quality-assured.
What I Learned About Multi-Agent Systems
“Don't build agents that try to do everything. Build agents that do one thing extremely well — and orchestrate them into a team.”
Agent Chronicles was my proof that multi-agent systems aren't just a buzzword. Agents are only as good as their role definition. An agent with a clear role, defined context, and measurable quality criteria delivers results indistinguishable from human work.
— Philipp
Multi-agent systems for your organization? Let's talk about orchestration and quality.
Book a consultation