Founder, Arda AI · Physical economy · Real-world execution

Raza Rafiq

I'm building Arda to make the physical world executable.

AI made intelligence abundant. The next bottleneck is execution — housing, energy, infrastructure, logistics, construction, repairs, and the messy work that keeps the real world moving.

I didn't arrive at this from a whitepaper. I arrived from job sites, capital stacks, deep-tech construction, and the operating businesses I've helped finance, build, and acquire — from the repeated experience of watching sophisticated plans break against messy execution.

Before the physical world can become autonomous, it has to become legible.

Arda is building the operating layer for that transition, starting with Goose.

Read the thesis → Current work

San Francisco · Toronto

Thesis

Abundance is bottlenecked by the physical world.

The last decade made intelligence abundant.

The next decade has to make execution abundant.

AI can write, reason, code, and plan. But human progress still depends on the physical world: homes, factories, roads, energy, materials, logistics, repairs, hospitals, farms, data centers, construction sites, and the people who keep them running.

That world does not move at the speed of software.

It moves through permits, crews, vendors, inspections, handoffs, site conditions, customer calls, material delays, field judgment, and thousands of decisions made under constraint.

This is where abundance gets stuck.

The physical world is not short on ambition. It is short on coordination.

Most software records what happened after the fact. Most AI still lacks the context, permissions, state, and audit trail to act safely in the real world.

What's missing isn't intelligence. It's the operating fabric that turns plans into action.

That gap is where I work.

Current Work

My current work is Arda.

Arda

/ˈɑːr-də/

noun

The world; the realm where action happens.

Borrowed from Tolkien's legendarium, where Arda is the created world. Here, it points to the practical ambition: building for the world that exists beyond the screen.

Company · Founder

Arda Industries

Arda turns the messiest parts of physical work — calls, jobs, decisions, exceptions — into structured state.

Most operating businesses run on tribal knowledge, fragmented tools, and whatever the owner remembers from yesterday. Arda is building the layer underneath: a shared operational truth that humans, AI agents, and eventually machines can act on together.

The goal isn't software for its own sake. It's more capacity — to build, operate, maintain, and repair the things people need.

From → to

Messy intake

Calls · texts · jobs · exceptions

Structured work

Clean job records · decisions · next actions

Shared operating truth

Humans · agents · machines acting from the same context

First product 2025

Goose

Goose is the first wedge into that operating layer. It starts in construction and service operations, where coordination is the daily failure mode — calls, quotes, jobs, schedules, dispatch decisions, approvals, exceptions, follow-ups. Goose turns that mess into structured, machine-readable work.

Not an AI receptionist. Not a dashboard. The first piece of an operating system for physical work.

Status
Design-partner stage
Active
2025 — Present
Get in touch

Operating Thesis

Four convictions shaping the work.

These are the patterns I kept seeing across capital, real estate, construction, and operations.

01 Conviction

Abundance depends on physical execution.

More intelligence does not automatically create more housing, infrastructure, energy, logistics capacity, or repair capacity. The bottleneck is turning plans into coordinated action in the real world.

02 Conviction

Physical work is coordination-constrained.

The barrier is not only labor, capital, or software. It is the handoff between people, materials, time, approvals, vendors, customers, jobs, and decisions.

03 Conviction

Tribal knowledge is operational infrastructure.

What people call “tribal knowledge” is often the company's real operating system. It determines margin, quality, speed, trust, escalation, and whether the plan survives contact with reality.

04 Conviction

AI needs legibility before autonomy.

Agents cannot safely operate physical workflows from loose context alone. They need state, permissions, audit trails, constraints, escalation paths, and decision history.

Journey

A career built inside the physical world.

Capital, real estate development, deep-tech construction, acquisitions, company building. Different chapters, same throughline: how the physical world gets financed, built, delayed, operated, and improved.

01

2016 — 2021

Altree Developments

Real estate development · capital formation

Worked across private equity, capital formation, and real estate development. Helped scale a platform to roughly 3,000 units of development assets, with exposure to $3B+ of pipeline.

Learned ·How physical assets are financed, underwritten, sold, delayed, de-risked, and ultimately shaped by coordination.

02

2021 — 2023

Mighty Buildings

Deep-tech construction · industrialized housing

Worked inside an ambitious attempt to industrialize homebuilding through robotics, materials science, modular manufacturing, and prefab construction.

Learned ·That construction is not just a technology problem. Even frontier technology runs into sequencing, permitting, supply chain, labor, site conditions, and execution.

03

2023 — 2025

Genos Group

Acquisitions · advisory · operating companies

Built a private holding and operating vehicle focused on acquisitions, advisory, and company-building in physical-world businesses.

Learned ·That the real operating system of many businesses lives in calls, vendors, hiring, follow-ups, quoting, scheduling, cash flow, owner judgment, and informal handoffs.

04

2025 — Present

Arda

Founder · physical-world coordination

Building software for real-world coordination, starting with Goose.

Now ·The current expression of a longer obsession: making physical work easier to understand, coordinate, and execute.

Working Notes

Notes on the physical world.

Working notes on abundance, execution, AI, coordination, and the systems that shape the real world.

01

Abundance Is Bottlenecked by the Physical World

Why intelligence alone does not create housing, infrastructure, energy, logistics, or repair capacity.

02

The Coordination Layer

Why the next great platforms won't be systems of record. They'll be systems of action.

03

Why AI Needs the Physical World

Agents become useful when they are grounded in real operations, constraints, permissions, and outcomes.

04

Tribal Knowledge Is Infrastructure

What people dismiss as “tribal knowledge” often determines speed, margin, quality, and trust.

05

Construction Is Manufacturing Without a Factory

A way to understand why sequencing, labor, materials, permitting, and site conditions make construction so hard to systematize.

06

Operational Judgment Is the New Data

The most valuable signal in many businesses is what the best operator notices, decides, and corrects under pressure.

07

The Company as a System of Action

What changes when a company is designed around decisions, permissions, actions, and feedback loops.

Work With Me

For people building where software meets reality.

I want to meet operators, engineers, investors, and builders working where AI meets the real world — construction, infrastructure, energy, logistics, field service, and the operating companies that hold them together.

If you believe the next frontier of AI is not just digital labor but coordinated execution in the real world, reach out.

Reach out →