Building Shippy
How Blank Metal is Reimagining Professional Services with AI (and Vercel)
We recently sat down with the Vercel team — including Alex Hawley and Arman Nijadiyan — to walk through Shippy, our internal AI platform that’s quietly reshaping how Blank Metal runs. What started as a prototype in April has become a live operating system for our business, built and deployed on Vercel, and now part of our daily client work.
The Problem: Professional Services are Broken by Design
We kicked off the discussion with a blunt truth: most consulting firms are structurally inefficient. Operating expenses regularly hit 35–50% of revenue, and nearly half of client fees are consumed by project management overhead. Clients hate that equation — and rightfully so.
From day one, Blank Metal set out to build a new model. What if we could use AI to:
Reduce OPEX to a fraction of traditional firms
Eliminate administrative waste (timesheets, reports, resource management)
Deliver higher-value work without drowning in a sea of SaaS licenses
That’s the thesis behind Shippy — an AI-driven operations layer for a professional services firm.
What Shippy Does
Shippy is built on a simple idea: every consulting workflow, from sales to delivery, can be automated or augmented with AI. The system is organized around four architectural layers:
Ambient Capture – voice, data, and context are automatically captured across tools and systems.
Intelligence Core – a unified data platform that ingests structured and unstructured data, forming the foundation for analysis and chat-based interaction.
Custom Web Apps – lightweight apps that let anyone at Blank Metal build and deploy tools without deep engineering experience.
Agent Suite – AI agents that take on real operational work: resourcing, forecasting, client prep, and project management.
We think of it as the modern architecture for an AI-native company, where software is the connective tissue for everything a services firm does.
From Prospect to Delivery: The Agent Roadmap
The Shippy roadmap maps directly to the client lifecycle:
Prospector: AI-driven client research and fit scoring before sales meetings.
Scope & Pricing Copilot: Generates accurate project scopes and price models in hours, not weeks.
Resourcer: Allocates and optimizes staffing automatically, adjusting for vacation schedules and utilization.
AutoPM: Keeps projects moving — updating JIRA, surfacing risks, and automating status reporting.
Timekeeper: Eliminates timesheets entirely.
Deliverable Orchestrator: Runs automated “pull requests” for deliverables, ensuring quality and consistency.
Guardian & Forecaster: Manage revenue realization and predict business performance in real time.
On top of these sits Shippy Chat, built and deployed on Vercel using the AI SDK. It gives Blank Metal leaders a conversational interface to query projects, resources, and forecasts — a single chat that knows the state of the business.
Built Fast, Built Differently
Shippy isn’t vaporware. It’s live, and the speed at which it was built has become part of the story.
The Resource Management Tool that took two years to build at a previous firm was rebuilt in one month.
The first version of the Pricing Engine was deployed in days.
RFP responses that used to take 7-10 days now take 36 hours.
The first version of Shippy Chat was made in 2 days.
That’s the power of pairing AI-assisted development (v0 and Vercel) with human judgment and lean process. Every engineer at Blank Metal spends their first 30 days contributing to Shippy, ensuring the platform keeps evolving with the company.
The Vercel Advantage: Infrastructure That Moves at the Speed of Ideas
Shippy’s rapid development has been accelerated by Vercel’s infrastructure serving as our foundation. While the platform enables the full scope of our vision, two core capabilities have been game-changers:
Preview Environments for Every Pull Request
Every PR automatically spins up a staging environment, letting us test features in isolation before they hit production. This eliminates the “works on my machine” problem and dramatically accelerates our feedback cycles. When you’re building agents that touch real business operations, being able to validate changes in a live-like environment before deployment isn’t just convenient, it’s essential.Deployment Simplicity Across the Team
Vercel’s deployment model has democratized our ability to ship. Our sales leader, operations lead, and engineers can all vibe-code a tool using v0, deploy it instantly, and have it running in production within hours. This cross-functional capability is critical: Shippy isn’t just an engineering project, it’s how the entire company operates. When non-engineers can contribute working tools, the pace of innovation multiplies.
The Path Forward
As we consolidate systems and scale autonomous agents, Vercel’s new Workflow Development Kit has fundamentally changed how we think about agentic tasks. The use workflow directive makes critical multi-step generative AI workflows durable—they can pause for hours or days without consuming resources, survive deployments and failures, and resume exactly where they left off.
This allows us to build more advanced agents that can delay notifications and tasks intelligently—send a status update, wait, then automatically follow up or send a notification based on what happened. Combined with the AI SDK’s agentic capabilities, we only pay for the compute we actually use while building truly autonomous operations.
What Happens Next
The next phase is about scale. The team is consolidating all systems — Slack, email, project tools, analytics — into a shared data lake to unlock cross-project visibility. From there, the goal is to start shipping fully autonomous agents built on Vercel infrastructure.
Clients don’t buy Shippy, they experience it. Shippy joins meetings, schedules follow-ups, and quietly handles the operations work no one loves to do. And once clients see it in action, they often start asking the same question: “Can you build us our version of Shippy?”
The Bigger Picture
Shippy isn’t just a product. It’s a way of thinking and behaving as a team so that we take advantage of all that AI has to offer. It’s the blueprint for how professional services firms can reinvent themselves with AI at the core: faster, leaner, and with radically less waste.
As our COO puts it, “If we do this right, we won’t just run a better firm. We’ll redefine what a consulting company looks like.”



