Building at the Bleeding Edge:
Zero-Code Product at Scale

Dan Shipper

Co-founder & CEO, Every

2025

The Company

Every: 15 People, 4 Products, Zero Code

"No one is manually coding anymore. Organizations like ours, people who are playing at the edge, we're doing things that in three years, everybody else is going to be doing."

4 products shipped with 15 people
Daily newsletter + publishing arm
Consulting that teaches AI best practices
Head of AI Operations role (new organizational function)
Zero handwritten code from engineers

Framework

The Every Operating Model

team size

shipped products

Spec → Prompt → Agent: Team writes requirements in plain language
Code Review Only: Developers review agent output, never write code
Agents Have Personality: Multiple Claude instances + Friday, Charlie agents for different perspectives
Spiral Tool: Internal low-code platform powered by AI agents
Fast Iteration: Small teams iterate at unprecedented speed

The workflow

Product spec → prompt engineering → agents write code → humans review quality → iterate. Zero manual coding required after initial setup.

New capability

Every team member is learning to manage AI agents instead of managing people. This is the skill that scales a 15-person team.

The Acceleration

How AI Amplifies Human Growth

Alex Duffy case study: Year of progress in 2 months by recording Dan's feedback into prompts, never repeating a mistake
Natasha's acceleration: Learned coding in AI era from day 1, now out-paces pre-AI engineers
Claude Code for everyone: Command-line interface that does autonomous work for 20-30 minutes unattended
The leash lengthens: AI capability measured by how long you can let it work unsupervised

The prompt library

Codify your best practices into reusable prompts. Teaching → recording → automating is the new mentorship model.

The allocation economy

Management skills will become cheap and ubiquitous. Everyone will manage AI agents. This is the job that scales at 15 people.

Playbook

Operating at the Edge

Hire generalists with multidimensional talent: Kieran (baker → composer → CTO → engineer)
Use multiple agents: Different agents have different personalities and perspectives
Codify knowledge in prompts: Turn feedback loops into repeatable templates
Let young talent grow fast: 20-year-olds with ChatGPT can match senior-level work
Measure progress by agent autonomy: How long can you leave it running?

Dan's insightWhenever I see a kid with ChatGPT, holy shit, they're going to go so much faster than any person I've worked with.

Contrarian Takes

AI Myths Dan Disagrees With

✗AI will destroy entry-level jobsINSTEAD →✓ Entry-level workers with AI move at senior speeds, skipping years of traditional climbing

✗Claude Code is only for engineersINSTEAD →✓ It's the most underrated tool for non-technical people processing large datasets autonomously

✗AGI is unknowable or unmeasurableINSTEAD →✓ AGI = when it's profitable to run agents 24/7 indefinitely without human supervision

✗AI will offshore American jobsINSTEAD →✓ Cheap AI intelligence reshores jobs by making US labor affordable and productive again

0:00