Since 2008, Zivtech has had one mission: remove the fear of using technological tools and illuminate the true potential of technologies to help the organizations we work with. We've done that through open source software, especially Drupal, and especially through building lasting partnerships with our mission-driven clients.
Now we're doing it with AI. Specifically, with Anthropic's Claude.
Zivtech is becoming a Claude Consultancy.
What Being a Claude Consultancy Actually Means
To get the most important part out of the way: this doesn't mean we're abandoning open source or Drupal. It simply means that we've fully embraced AI, which we believe means we're about to get a whole lot better, faster, and more ambitious.
For us, this transformation has two parts.
First, we're restructuring how we work internally. Claude Code is now our primary work environment for development — but we are also mandating it for project management, discovery/design, sales, and business processes. Everyone at Zivtech has to use Claude in their daily work.
Every Skills file we write, every workflow we refine makes the next one better and faster. We're not just getting better at this — we're getting better at getting better.
Second, we're offering Claude services to our clients, with services like training & enablement, custom Claude Skills/Plugin development, multi-agent platforms, regulatory/control/assurance/fidelity monitoring, and workflow automation. We can help navigate the landscape of AI integrations and possibilities and, most importantly, we can help you start to retool your staff to work at the speed of AI. We're learning something new every day, we want to share what's working, and we want to help others start their AI journeys in ways that are both empowering and safe.
Claude — But Not Only Claude
Being a "Claude Consultancy" doesn't mean we only use Claude. We work with the full range of AI tools, including Gemini, OpenAI, and others, and we'll help clients with whatever tools make sense for their situation. In my daily workflow I regularly work across Claude, ChatGPT, Codex, and Gemini, using each for the different tasks that research and benchmarking suggests they are best suited for.
We've also developed a real fondness for the agentic tools that orchestrate these large language models (LLMs). Tools like Perplexity for research, n8n or Activepieces for workflow automation, and Manus for autonomous agents are becoming essential parts of how we work.
Two tools have become especially central to how we orchestrate all of this: oh-my-claudecode, a multi-agent orchestration layer for Claude Code, and Spec Kitty, our spec-driven development system. Both are worth their own posts — and those are coming.
Why Claude as Our Foundation?
So why are we aligning ourselves so closely with Claude? We've worked with most of the major AI tools. We chose Claude as our primary platform for a few reasons.
First, it's where we've gone deepest, and where we have found the most value. Claude Code is our daily driver for development, but I am also incredibly excited by — yes, even addicted to — the transformative way it can help solve business issues. Claude is exceptionally good at understanding context, and it's great at reasoning through complex problems alongside experts.
Second, Anthropic's business philosophy closely aligns with our own. From their approach to safety and transparency, to their research focus, to their education efforts, to their pricing: we feel that Anthropic aligns with our values. Anthropic's thoughtful focus on building AI responsibly resonates deeply with how we've always approached technology.
We're still pretty tool-agnostic when it comes to solving client problems — we use what works, and many of the LLMs can support the same workflows we've standardized on. So while we are Claude-first, we are not (yet) Claude-only.
Always Learning
Given the speed of these advancements, we are, and will be for the foreseeable future, in active discovery and learning mode.
I haven't been this excited about technology since the Drupal 4.7/5 days. Back then it felt like the ground was shifting under our feet in the best possible way — empowering motivated people to bring real change and advancements to all sorts of people and organizations. New features brought amazing new capabilities, and the incredible Drupal community was figuring things out together in real time. The possibilities seemed endless.
That's where I feel we are now with AI, except at a level I never imagined possible. Every week brings new tools, new capabilities, new ways of working. I'm not claiming to have all the answers — given the speed of change expertise is measured in months and years instead of decades. We're learning alongside everyone else, but we're learning fast, and we're learning by doing. To say that I'm excited by all of this is an incredible understatement: I have never felt more keenly that a technology had the potential to change the world for both better and worse, and I've never been more motivated to ensure we end up on the better side of that equation. I'll be writing a lot more about this over the next few months, but Claude also dramatically changes the landscape of potential tools and expertises that we can bring to bear on problem spaces that were off limits due to time and cost constraints.
Open Source Isn't Going Anywhere — It's Getting Stronger
Becoming a Claude Consultancy doesn't mean we're walking away from our open source roots. Our team has been active in the Drupal community for nearly two decades. We advance the open source platforms on which we work by contributing our time, code, and knowledge. In return, we stay on the cutting edge of technology and our work benefits from consistent peer review.
Our belief is that AI is going to let us contribute more. When a developer can move through routine tasks faster, they have more time for the work that matters, including upstream contributions. When we can explore more solutions in less time, we can share those solutions with the community. The backlog of "things we'd contribute if we had the time" is about to get a lot shorter.
We anticipate dramatically increasing our contributions to the Drupal project. Not despite embracing AI, but because of it. I've already submitted patches to drupal.org for the first time in over 15 years. That's not a coincidence.
We want to bring that same open source ethos to AI itself. We're looking for ways to open up what we learn and build, to share our discoveries, our Skills, our workflows with the broader community. If you check out the Zivtech Github account you'll see some of the ways we're thinking about this new work, and you'll see much more coming, and at an ever increasing speed, over the next year:
- Joyus-AI — An open-source platform for deploying AI agents that treats your organization's knowledge — your standards, your voice, your workflows — as a first-class part of the AI rather than an afterthought. The core idea: every organization should be able to shape how AI works for them, not just accept the defaults. We're building this in the open because governance and customization tools should be available to everyone, not just the biggest enterprises.
- Harsh-Critic — A code review skill for Claude Code that uses a structured investigation protocol to find what's wrong and what's missing — the problems reviewers typically skip because they were never prompted to look. Standard AI reviews scan what's there; this one actively hunts for what isn't. It's the foundation our other review tools are built on.
- Drupal-Planner — A planning skill that designs Drupal implementations before the first line of code is written: entity structures, permission models, cache strategies, migration paths. Drupal architecture decisions are expensive to reverse, so we made a tool that forces the hard thinking upfront. Works as a pair with Drupal-Critic — plan first, then review.
- Drupal-Critic — A Drupal-specific code review skill that knows what generic reviewers don't: cache tag invalidation, config export safety, contrib-first decisions, update hook idempotency. It builds on Harsh-Critic's protocol and adds the domain knowledge that comes from nearly two decades of Drupal work. This is us encoding what our best developers know and making it available to everyone.
- React-Critic — A review suite covering React, Next.js, and React Native/Expo, built the same way as Drupal-Critic: structured investigation, evidence requirements for critical findings, and specialist perspectives that know the framework's specific failure modes. Includes a router that dispatches to the right critic based on what it detects in your code.
If you're building in public, sharing what you learn, trying to figure this out for the benefit of everyone: we want to collaborate.
What This Means for Our Clients
We're bringing all of this to clients: training and enablement for teams figuring out where to start, custom Claude Skills built around how your organization actually works, and integrations that wire AI into your existing systems. And if you're not sure what you need yet, that's fine. That's often where we start.
This applies beyond software shops, too. I use Claude Skills as a live intelligence layer for Milk Jawn (an ice cream company I own in Philadelphia), running financial analysis, staffing decisions, and hiring. If the work involves judgment and information, Claude can probably help with it.
Part of this work is also about governance — helping clients deploy AI in ways that protect them from the real risks these tools bring. That's part of what led us to start building Joyus.ai, an open-source platform for responsible AI deployment. Much more on that soon.
The Honest Part
We won't pretend this transition is simple. Some of our team members jumped in enthusiastically. Others were skeptical at first. All of them are learning and adapting.
What we know is this: the organizations that figure out AI-augmented work now will have a real advantage. The teams that learn to work with these tools will build faster, explore more possibilities, and deliver better outcomes. We've seen it in our own work already.
This is what Zivtech has always done. We find the technologies that matter, we get good at them, and we help our clients do the same. Open source. Drupal. DevOps. Now AI.
AI is the biggest shift in how software gets built since we started this company. We're not sitting it out.
Let's Talk
If you're thinking about how AI fits into your organization, whether that's training your team, building custom tools, or just figuring out where to start, we'd love to hear from you. We're learning something new every day, and we're ready to share what we're finding.
And if you're part of the Drupal community, or any community of civic-minded technologists trying to figure this out together, watch this space. We're about to have a lot more to contribute, and we want to do it in the open.