Arslan Rehmani
Return to Case Studies
Textile Manufacturing•Timeline: 8 weeks from initial floor audit to full factory cutover
Live production deployment

TextileMode ERP

Designed and deployed a full-stack custom ERP covering yarn lot intake, machine allocation scheduling, on-loom greige inspection, piece-rate shift payroll, and executive balance sheets. Deployed live at erp.textilemode.com.

Next.js App RouterPostgreSQLTailwind CSSVercel StandaloneDocker

01. Client Context

TextileMode is a commercial textile mill managing 27 active industrial looms producing greige and finished fabric orders.

02. The Problem in the Client's Words

"We have 27 looms running 24 hours a day. Every shift change, supervisors write piece numbers and yarn weight on paper registers. By Thursday, no one knows which loom is running which lot, our stock of 30/1 yarn is off by hundreds of pounds, and payroll has to recalculate piece-rates by hand on Sunday night."

03. Constraints & Realities

  • Zero production stoppage: factory runs 24/7 in 12-hour shifts.
  • Supervisors with minimal computer training using rugged floor tablets.
  • Harsh industrial environment with intermittent Wi-Fi drops.
  • Audited double-entry ledger required for commercial bank facility audits.

04. Solution & Architecture Blueprint

Data flow showing intake, deterministic validation barriers, and permanent double-entry audit logging:

Architecture Blueprint: TextileMode Production ERPZero-Loss Data Pipeline
01. Floor IntakeRugged floor tabletsShift supervisor PINTouch-numeric input27 Looms02. Validation GuardRange & density checkLoom allocation stateDuplicate lot rejectZero-Corrupt Gate03. Core LedgerPostgres RLS engineDouble-entry journalsImmutable audit logEncrypted at rest04. Close11 reportsReal-time P&LPiece wagesAudited close

05. Data & Security Design

Access Rules (RLS)

Postgres Row-Level Security: shift supervisors can only insert inspection and run logs for their active shift; accountants have read-only access to floor data; only the managing director can authorize ledger closures.

Audit Trail

Every production record change logs actor ID, timestamp, before/after JSON hash, and machine IP. Deleted records are soft-flagged and trigger immediate email notifications.

Staging & Validation

Input validation guards on tablet entries: yarn weights must be within 2.5% of theoretical density; impossible meterage counts throw immediate visual validation blocks before insert.

06. Measured Results (Truth Gate Verified)

Backed by facts.json
27 looms
Looms Managed in Real Time

Active production scheduling dashboard and shift tracking

Period: Continuous since launch

40+ hours
Weekly Clerical Work Eliminated

Audit of previous manual shift register transcription across 5 clerical roles

Period: Measured across first 90 days

10+ months
Continuous Production Uptime

Live deployment uptime and daily transaction log records

Period: 10+ months live operation

07. What Went Wrong & What I Changed

Candid Engineering Retrospective

During the first week of live cutover, night-shift supervisors struggled with touchscreen dropdowns when wearing work gloves. They reverted to paper logs for two nights until we rebuilt the entry screen with large numeric tap pads and single-stroke validation buttons.

09. What I'd Do Differently

I would have spent the first three days directly on the night shift observing tablet ergonomics under dim factory lighting before writing the first UI component.

Need a system engineered to this standard?

Book a 20-minute discovery call to review your operational bottlenecks.

Book a 20-minute discovery call