Tokenless

The AI-native startup: 5 products, 7-figure revenue, 100% AI-written code. | Dan Shipper (Every)

Dan Shipper, CEO of Every, shares how his 15-person team operates on the bleeding edge of AI, shipping products without writing code, using a team of specialized AI agents, and pioneering new AI-first workflows. This summary covers Every's operational playbook, their AI stack, and Dan's predictions on how AI will reshape jobs, skills, and companies.

5 min read
The AI-native startup: 5 products, 7-figure revenue, 100% AI-written code. | Dan Shipper (Every)
source: https://www.youtube.com/watch?v=crMrVozp_h8

Dan Shipper’s company, Every, operates at the frontier of AI-native companies. With a small team of 15, they publish a daily newsletter, ship multiple software products, and run a consulting arm by deeply integrating AI into every facet of their operations. Their engineers have moved from writing code to managing an arsenal of AI agents, a model that suggests a future where job roles are fluid and everyone becomes a manager of AI.

The AI-First Operating Model

At Every, the operational philosophy is built around maximizing leverage with AI. This has led to the creation of new roles and workflows that may become standard in the future.

  • Head of AI Operations: Every employs a “Head of AI Operations” whose sole focus is to identify repetitive tasks across the company and build AI-powered workflows and prompts to automate them. This dedicated role ensures that the burden of process innovation doesn’t fall on team members focused on execution, increasing the likelihood of adoption.

  • Compounding Engineering: The engineering team operates on the principle of “compounding engineering,” where every task should make the next similar task easier. Instead of repeatedly writing detailed Product Requirements Documents (PRDs) from scratch, they build and refine prompts that can convert rambling thoughts into structured PRDs for AI agents. This work is stored in a shared GitHub repository of prompts.

  • Engineers as AI Managers: The product team at Every does not manually write code. Their workflow has shifted to:

    1. Crafting highly detailed PRDs, often using their “compounding engineering” prompts.
    2. Feeding these PRDs to AI coding agents like Claude Code.
    3. Spending their time reviewing the generated code, managing the agents, and refining the output.

    “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.”

  • An “Avengers” Team of AI Agents: The team uses an arsenal of AI agents in parallel, treating each one like a specialist with unique strengths, tastes, and personalities. This includes using Claude, Codex, and other specialized agents like “Friday” and “Charlie” (which integrates directly into GitHub pull requests). This approach suggests a future with a diverse market of AI agents, not a single winner-take-all.

The Modern AI Stack

Every leverages a suite of frontier models and tools, choosing the right tool for the job.

  • ChatGPT (o3/GPT-4): The go-to for daily tasks, writing, and self-reflection, primarily due to its “Memory” feature which provides personalized and context-aware responses over time.
  • Claude Code: The workhorse for the engineering team. It’s an agentic command-line interface that can access the local file system, run for long periods, and even spawn sub-agents to complete complex coding tasks autonomously. Shipper’s hot take is that it’s the most underrated tool for non-programmers for tasks involving processing large amounts of local text files (e.g., analyzing meeting notes, studying writing styles from books).
  • Claude Opus 4: Praised for a unique capability no other model has: a “gut sense” for judging the quality and interestingness of writing. This unlocks use cases where the model can act as a judge, enabling it to critique its own work and iterate before showing it to a human.
  • Gemini: Used extensively within Every’s applications due to its power and cost-effectiveness.
  • Speech-to-Text Interfaces: Tools like Granola and internal equivalents are seen as the future, making it easier to capture and process information.

The Future of Work and Skills

Shipper believes AI will fundamentally change the labor market and the skills required to succeed.

  • The Allocation Economy: We are moving from a knowledge economy to an “allocation economy.” The most valuable skills will mirror those of a great manager: vision, taste, talent evaluation (of AIs), and knowing when to dive into the details versus when to delegate.
  • The Rise of the Generalist: As AI handles hyper-specialized tasks, it empowers individuals to work across multiple domains. This will allow small, agile teams of generalists to accomplish what previously required large, siloed organizations. Rigid job titles will blur as everyone manages a suite of AI tools.
  • Accelerated Learning: AI acts as a powerful mentor. Junior employees can progress dramatically faster by codifying feedback from senior members into prompts, ensuring they “never make the same mistake twice.” A new hire at Every made a year’s worth of progress in two months using this method.
  • Coding Knowledge Remains Vital: While engineers at Every don’t write code, they must understand it deeply to effectively manage agents and review their work. English is becoming a new layer on top of scripting languages, similar to how scripting languages were a layer on top of C. For the foreseeable future, knowing how to code will remain a significant advantage.

Playbook for Corporate AI Adoption

Based on Every’s consulting work, a clear pattern has emerged for companies that successfully integrate AI.

  1. CEO-Led Initiative: The single biggest predictor of success is whether the CEO personally uses AI tools like ChatGPT daily. This ensures they have an intuitive grasp of the technology’s capabilities and limitations, allowing them to drive excitement and set realistic expectations.
  2. The “AI-First” Memo: The CEO should send a company-wide memo declaring an AI-first strategy. Shipper highlights a powerful move: ending the memo with, “I wrote this email with ChatGPT, and you should too.”
  3. Create Internal Forums: Establish weekly meetings and email digests where employees can share successful prompts and use cases. This identifies and empowers the 10% of employees who are early adopters, allowing their knowledge to spread throughout the organization.

Share post

Stay In The Loop! Subscribe to Our Newsletter.

Get updates straight to your inbox. No spam, just useful content.

Recommended for You

No Priors Ep. 123 | With ReflectionAI Co-Founder and CEO Misha Laskin

No Priors Ep. 123 | With ReflectionAI Co-Founder and CEO Misha Laskin

Misha Laskin, co-founder of Reflection AI and former researcher at Google DeepMind, discusses the company's mission to build superhuman autonomous systems. He introduces Asimov, a code comprehension agent designed to solve the 80% of an engineer's time spent on understanding complex systems, rather than just code generation. Laskin delves into the intricacies of co-designing product and research, the critical role of customer-driven evaluations, the bottlenecks in scaling reinforcement learning (RL) — particularly the "reward problem" — and why he believes the future is one of "jagged superintelligence" emerging in specific, high-value domains like coding.

DeepMind's Pushmeet Kohli on AI's Scientific Revolution

DeepMind's Pushmeet Kohli on AI's Scientific Revolution

Pushmeet Kohli, head of AI for Science at DeepMind, discusses AlphaEvolve, an AI system that uses Large Language Models (LLMs) coupled with evolutionary search to discover novel, human-interpretable algorithms. He explains the architecture, from its predecessor FunSearch to the multi-agent "Co-scientist" system, and details breakthroughs in solving decades-old math problems and optimizing real-world systems like data center scheduling and chip design.