Project Weaver: The Journey So Far and What’s Next

When we first imagined Project Weaver, it was a big, audacious idea: Could we build an AI-powered engineer that truly supports the creation of serious, production-ready backends? Not a toy tool. Not a research prototype. Something that could sit alongside human engineers and help them ship real work with real impact.
Over the past several months, we’ve been building in public — documenting our thinking, sharing what we’ve learned, and inviting the community into a conversation about the future of AI-augmented software development. Below is a recap of that journey so far, with links to each of the blog posts that trace this evolution.

1. Why We’re Building Our Own AI-Powered Development Stack - and Why Now

Our first article laid the foundation by describing why we chose to build our own stack instead of relying entirely on existing tools. We explored the limitations of off-the-shelf AI coding assistants and explained our belief that a predictable, opinionated environment is key to reliability, safety, and real engineering value.
Read Blog 1 here:

2. Where Expertise Meets Innovation: The Leader Behind Project Weaver

Next, we introduced the leader guiding Weaver’s technical vision — blending deep engineering experience with a curiosity about how AI can best support software craft. This article gave context to our team’s background and how that expertise shapes the decisions and priorities in Project Weaver.
Read Blog 2 here:

3. Building in Public: Why We’re Sharing the Evolution of Project Weaver

From the outset, we committed to transparency. In this post, we talked about why we’re sharing our work publicly — the benefits of community feedback, the accountability it creates, and how it aligns with our mission to elevate not just our team but the broader engineering community as we explore AI-assisted workflows.
Read Blog 3 here:

4. Why We Chose NestJS for Project Weaver: Building an Opinionated AI Engineer for Serious Backends

One of the first major technical decisions we documented was our choice of tech stack. This article walks through why we selected NestJS as the backbone of Weaver’s output — a framework that balances developer ergonomics with scalable, maintainable architecture — and how that choice informs every specification and code generation pattern we build on top of it.
Read Blog 4 here:

5. Project Weaver: Building an AI Engineer in Three Phases

With vision and stack in place, the next step was to define a clear roadmap. This article introduced the three phases guiding Weaver’s development — from documenting workflows and best practices, to automation, and finally to project-level orchestration — with a timeline and rationale for each stage. This phased approach helps us focus, measure progress, and ensure each milestone delivers value internally and for the community.
Read Blog 5 here:

6. Spec-Driven Development: How AI Is Changing the Way We Think About Software Design

Most recently, we dove into how the rise of AI shifts what engineering work looks like. In this piece, we unpacked spec-driven development — why effective specifications are the real interface between human intent and AI execution, and how both human-authored and AI-generated specs are essential for reliable outcomes. This article represents a conceptual milestone: not just building an AI engineer, but understanding _ how_ to think about engineering in the age of AI.
Read Blog 6 here:

Where We Are - and What’s Ahead in the New Year So far, Project Weaver has been about laying the intellectual and architectural groundwork and testing:

• Choosing a stack we trust
• Defining phased development
• Exploring how AI changes engineering practice
• Sharing our reasoning, not just polished demos

Now, as we close out this year and look to the next, we are ready to shift more of the conversation toward technical depth, demos, testing app development, and results.
In the coming months, we plan to share:

Technical walkthroughs and live demos

Seeing Weaver in action, from plan generation to code output, and how it navigates real engineering scenarios.

How specifications translate to working code

Concrete examples of human specs and AI working specs producing reliable, consistent backend features.

Early metrics and learnings

What we’ve learned about accuracy, reliability, error modes, and how engineers interact with the system in real workflows.

Feedback loops from real users

We’ll open up early usage, work with some design partners, gather feedback, and iterate in public so you can follow the evolution as it unfolds.
Our journey is just beginning, but we’re excited to move from theory and design into _ practice and impact._ If you’ve been following along, thank you. If you’re just joining, you’re welcome -there’s never been a more exciting time to rethink how software gets built.
Stay tuned for more technical depth, honest reflection, and the first real demos of Weaver making engineering work more predictable, productive, and collaborative.
Let’s see what the future of engineering looks like, together.