Subtitle: A bespoke fashion business management system, built by the person who needs it
Started: February 2026
Status: Active
Tools & Materials: Next.js 14, Prisma, PostgreSQL, Tailwind CSS, shadcn/ui, TypeScript, Claude Code
Project Brief
FashionSuite is the software I'm building to run Cordon Bleuet, my one-person bespoke atelier. Nothing on the market fits a solo maker doing made-to-measure work — enterprise PLM tools are built for teams of fifty managing hundreds of SKUs, and generic project managers don't understand that a welt pocket has dependencies. So I'm building my own.
The core idea is GTD-inspired task management for garment production. The killer feature is "What Should I Do Now?" — a context-aware query that looks at where I am (cutting table, sewing machine, fitting room), how much time I have, and what energy I'm bringing, and tells me the most important thing to work on. Customer orders always come first. Development work fills the gaps.
The Vision
A system I open on my phone in the workshop and it tells me what to do. Not a dashboard I have to interpret — a direct answer. The task system understands that construction can't start until the pattern is confirmed, that the second fitting blocks the final hem, that the lining has to be ordered before I can start the jacket body. Real dependencies, real critical path analysis.
Beyond the day-to-day, I want this to manage two parallel pipelines: customer commissions (the committed, deadline-driven work) and new product development (the creative work that fills downtime and builds the product line over time). Every NPD project follows a Learn, Practice, Sell progression — you don't sell something you've only made once.
If it ever becomes something other people use, that's a separate conversation. Right now it's one maker building one tool because nobody else made the right thing.
Build Log
2026-02-20 — Taking Stock
I'm starting this build log partway through the project, so the first entry is a snapshot of where things stand rather than a story about what just happened.
FashionSuite has a working codebase. The foundation is solid — Next.js with App Router, Prisma ORM talking to PostgreSQL, Tailwind and shadcn/ui for the interface, NextAuth handling authentication. That's Sprint 1 from the MVP plan, and it's done.
Customer management works. I can create and edit customer profiles, store body measurements, search and filter. The design library is functional too — designs with status tracking, filtering, estimated production times. That covers the core entities everything else depends on.
Order management is where it gets interesting. Orders link customers to designs with line items, there's task auto-generation from design templates, status workflows, due date tracking, material linking, and payment status integration. The order detail view has dedicated cards for customer info, due dates, line items, notes, pricing adjustments, and status actions. That's Sprints 2 and most of 3.
The task system — the heart of the whole thing — has a context selector, energy and time filtering, task actions with start and complete buttons, and a filter bar. The "What Should I Do Now?" query exists, though I haven't battle-tested it in actual workshop conditions yet. Time tracking has a timer widget, manual entry, and order-level summaries. Materials have basic inventory and adjustment tracking. Payments have CRUD and status tracking.
There's even early work on things I didn't expect to reach yet: a Gantt chart component, PWA scaffolding with offline indicators and sync status, mobile navigation, production reports, and a settings API.
What I learned:
Building a feature list in code is not the same as having a working system. There's a gap between "the component renders" and "I trust this enough to manage a real commission with it." That gap is where the next phase of work lives. I'm still finding places to refine or features I did not think of that I need to add to improve the user experience.
What's next:
Use it to track my next school project, to see where it breaks, where it's slow, where I reach for paper instead of the screen. That feedback loop will tell me more than another month of feature development.
Technical Notes
Stack:
- Next.js 14 with App Router
- Prisma ORM with PostgreSQL
- Tailwind CSS + shadcn/ui component library
- NextAuth.js for authentication
- TypeScript throughout
- Jest for testing
Architecture:
- Dashboard layout with authenticated routes
- REST API routes under
/api/ - Context-based state management (TimerContext)
- Validation layer (
src/lib/validations/) - Separated component hierarchy: layout, UI primitives, domain components (customers, designs, orders, tasks, time-tracking, PWA)
Feature coverage by MVP sprint:
- Sprint 1 (Foundation): Complete — project setup, core schema, customer CRUD, design library
- Sprint 2 (Orders): Complete — order schema, creation flow, task generation, list/detail views
- Sprint 3 (GTD Tasks): Mostly complete — core query, task UI, status workflow, time tracking
- Sprint 4 (Materials/Payment): Complete — basic inventory, payment recording
- Sprint 5 (NPD): Not started
- Sprint 6 (Polish): Partially started — PWA scaffolding, mobile nav, offline indicators
Reflections
None yet. I'll write the first one after I've run a real order through the system. That's when I'll know if this tool actually changes how I work, or if it's just a very elaborate form of procrastination.
Project Status
Current phase: Post-Sprint 4 — preparing for real-world testing
Completion: ~65% of MVP (Sprints 1-4 complete, 5-6 remaining)
Last updated: 2026-02-20
Key dates:
- Project started: February 2026
- Sprints 1-4 code complete: February 2026
- First real order through system: TBD
- NPD pipeline (Sprint 5): TBD
- PWA polish (Sprint 6): TBD
This build log is a living document. It gets updated, not replaced. Read it top to bottom and you get the full story of how this thing got made.