Sailing into the unknown — Pickaxe Building in Public series

AI never sits still. And neither can we.

Clayton Christensen’s The Innovator’s Dilemma describes a pattern that has played out across industries for decades. Successful companies don’t fail because they stop innovating. They fail because they keep optimizing for the thing that made them successful, right up until the ground shifts beneath them. The most rational-sounding decision, keep doing what works, turns out to be the most dangerous one.

We think about this a lot at Pickaxe.

What Got Us Here

Since we started Pickaxe, our mission has been simple: give people the best technology to build, deploy, and monetize AI agents. That hasn’t changed. But the landscape around us looks almost nothing like it did when we began.

We built a great product. One that fits the RAG model perfectly. Pickaxe has helped consultants, coaches, creators, and business owners package their expertise into AI-powered tools and actually make money doing it. Real businesses have been built on top of our platform, and that matters deeply to us.

But here’s the tension we keep running into.

The New World of AI

The next generation of AI looks vastly different from the one we built for. Trying to bend the latest technology to fit inside a UI designed for legacy infrastructure is becoming more and more redundant.

This is the classic Christensen trap. The existing product works. Customers depend on it. Every incentive tells you to keep refining what you have. But the technology is moving in a direction that your current architecture was never designed for, and no amount of incremental improvement will close that gap.

Everyone is a developer now. People want more control. They want to write code, customize behavior, and leverage AI agents to help them build. The rise of agentic workflows means a more powerful API interface would let agents themselves co-build inside the platform.

We’ve seen this firsthand. When we build new OpenClaw instances for ourselves, the process looks nothing like going through the Builder UI. It’s us chatting through the CI/CD pipeline. That’s the way we want to build our product, and it’s the way our users increasingly want to build theirs.

There’s another shift we can’t ignore. More and more companies are building appless apps. Tools that don’t really have a UI at all. They just integrate natively with the messaging platforms people already use: iMessage, WhatsApp, Slack.

So we’re at a crossroads.

We want to keep building on the cutting edge. But we also don’t want to break what our best users have already built their businesses around. Forcing large architectural changes onto an existing product that people depend on isn’t fair to them.

The answer, we believe, is to build something entirely new.

What We’re Building

The new product will be closely tied to Pickaxe but separate enough to give us the freedom to make bold changes without disrupting current users. Think of it as a much simpler, much more powerful version of what Pickaxe offers today.

Here’s the vision.

You create powerful OpenClaw agents for yourself and your users at scale. You manage and monetize them. Your clients can tailor the agents to fit their specific needs, too.

Instead of one portal housing many disconnected agents, you sell a single unified agent connected to any skills you want to add. You can offer premium skills or add-ons to enhance it. One agent, endlessly extensible.

Deployments get simpler too. We’re focusing on messaging-native deployment types: iMessage, WhatsApp, Slack. The places where people already spend their time and where chatting with an AI assistant feels most natural.

The platform will be developer and agent-friendly. More adjustable through code. More room for AI agents to help you build. One of the hardest unsolved problems in this space is deploying, managing, and updating agents that are themselves being customized by end users. That’s a technically challenging puzzle, and we want to help solve it not just for consultants and coaches but for startups and developers building next-generation AI tools.

Building in Public

As we embark on this journey, we want to document it and share it with you.

We’re living in a wild, fast-paced world where every week brings new learnings and forces important business decisions. It has been extremely fun and exciting for us. And honestly, we felt like we should be writing it down.

Not just for our own sanity, but because we genuinely think that for better or worse, it’s going to be a great story.

So we’ll be building in public. Sharing our highs and lows. All the challenges and all the learnings.

We’re excited to have you join us for this next adventure.

Related Articles

Building in public - why we wrote documentation before code at Pickaxe
Building in Public

Why We Wrote the Docs Before the Code

When we decided to rebuild from scratch, the first thing we did wasn't write code. It was write documentation. Here's why that decision changed everything about how we build.

April 07, 2026Read more
How to sell AI agents to local businesses — illustrated village scene with a small adventurer delivering a glowing companion to a cozy shop
Strategy & Business

How to Sell AI Agents to Local Businesses ($300–$1,500/Month Per Client)

The exact playbook for selling AI agents to local businesses — from picking your niche and pricing at $300–$1,500/month to landing clients and scaling to a full agency.

May 11, 2026Read more
How to start an AI agent agency - futuristic landscape representing the AI agency opportunity in 2026
Strategy & Business

How to Start an AI Agent Agency: The Complete 2026 Playbook

The step-by-step playbook for starting an AI agent agency in 2026 — from picking your niche and pricing your services to landing clients and scaling beyond solo.

April 14, 2026Read more
Hermes Agent vs OpenClaw comparison - two approaches to AI agents in 2026
Comparisons & Reviews

Hermes Agent vs OpenClaw: Two Paths to Your Personal Super-Agent Infrastructure

Hermes and OpenClaw are the two leading platforms for building personal AI agent infrastructure in 2026. We compare their architectures, memory systems, and self-improvement capabilities to help you pick the right foundation for your always-on AI.

April 13, 2026Read more
How to build an AI agent for client onboarding step by step guide
Guides & Tutorials

How to Build an AI Agent for Client Onboarding (Step-by-Step)

A practical, step-by-step guide to building an AI-powered client onboarding agent using Pickaxe — from defining the workflow to deploying it on your website.

April 10, 2026Read more
Top AI platforms in 2026 - comparison and review of the best platforms
Comparisons & Reviews

Top AI Platforms in 2026: The 15 Best Platforms I've Actually Tested

I tested 15 of the top AI platforms in 2026 — from ChatGPT and Claude to Pickaxe and OpenClaw. Here's my honest breakdown of what each one does best, pricing, pros and cons, and who should use them.

March 31, 2026Read more