Skip to main content
Build in Public8 min read

9 Years in Supply Chain Digitization to AI Builder: Shipping 7 Real Products Solo

Beyond empty AI wrappers: deliver working code as the sole standard. A retrospective on bridging architecture, coding, database migrations, and deployment with AI agents.

#AI Builder#Build in Public#Solo Developer#Full-Stack Engineering#AI Coding Agents#Supply Chain Digitization

Why Code and Working Runtime Are the Sole Delivery Standard

Having spent 7 years in the digitization of apparel, textile, and cross-border supply chains, followed by 2 years in government-enterprise complex systems, I developed an unbreakable operating premise: Any discussion not aimed at verifiable production execution is a waste of time.

Traditional software cycles are weighed down by coordination drag: PRD reviews, UI iterations, frontend-backend API alignment, DB approvals, QA cycles, and release windows. What should be a compact utility often consumes weeks of organizational overhead.

When LLMs and AI Coding Agents reached production viability, individual leverage multiplied hundredfold.

However, superficial hype followed: tens of thousands of shallow AI wrappers flooded the market, collapsing under real database constraints, network latency, and edge cases.

At Shadow Lab, we instituted one immutable principle:

Deliver working code as the sole standard. Build focused tools that resolve acute friction points in concrete workflows.

The 7-Product Matrix: Solving 7 Concrete Friction Points

People often ask: "How can a solo builder maintain and ship 7 distinct products simultaneously?"

The key is not burnout hours, but strict constitutional automation and hyper-focused scope:

  1. Quick flomo (Chrome Extension): Resolves the "bookmark graveyard" problem with inline clipping, AI summarization, and academic citations directly into flomo.
  2. Shadow Card (Web App): Transforms paper cards and ephemeral links into an enduring 4K dynamic geek profile.
  3. Shadow Snap (Client-Side Canvas): Eliminates bulky editing software for stitching cinematic subtitle long-images entirely inside the browser.
  4. Shadow Vision (Open-Source MCP Service): Equips coding agents like Cursor and Claude with multimodal vision, screen OCR, and UI-to-code reconstruction.
  5. Shadow Size (Apparel Sizing Engine): Leverages 7 years of garment engineering to provide zero-photo size guidance, attacking cross-border return rates.
  6. Shadow Mate (Family Habit PWA): Replaces screen addiction and nagging with offline-first habit loops and tangible pen-and-paper milestones.
  7. Shadow Shift (Open-Source CLI): Rescues years of locked WizNote archives into Obsidian with reconstructed bidirectional WikiLinks and attachments.

The Architectural Foundation of Solo Leverage

Shipping reliably across 7 tools rests upon three disciplined engineering pillars:

1. Canonical Source of Truth

Every product attribute—pricing, versioning, URLs, and metadata—derives strictly from lib/products.tsx. No drifting duplicate lists exist across scripts or UI pages.

2. Constitutional Database Control Plane

Shared Supabase infrastructure enforces audited release governance. Product repos submit proposals; migration execution follows immutable serial ledgers and row-level security tests.

3. Self-Healing Atomic Deployment

Vercel auto-deployments are deactivated in favor of deterministic scripts that verify Node runtimes, lock team scopes, build, and run dual-domain health checks before promotion.


The Path Forward: From Utilities to Commercial Value

Solo shipping is not just a technical showcase—it is an engine for authentic commercial value. Micro-tools attract targeted long-tail traffic and passionate builders, while vertical engines like Shadow Size directly tackle multi-billion-dollar reverse logistics friction for global e-commerce.

Turning an idea into working software is step one; solving real troubles that generate sustainable revenue is the builder's ultimate frontier.

All Field Notes