All Case Studies
Product Engineering Mar 2025

How Trellis Cut Time-to-Market by 11 Weeks

A Series A logistics startup needed a shipper portal rebuilt from scratch — fast. We delivered a production-grade Next.js platform with a serverless AWS backend in 9 weeks, 11 weeks ahead of their original estimate.

Logistics warehouse with bright overhead lights, workers operating forklifts, clean industrial floor, high ceilings, neutral tones

11 wks

Faster time-to-market

9 wks

Actual delivery

99.98%

Uptime at launch

Faster page loads

The Challenge

A portal that was costing them deals.

Trellis had raised a $4M Series A with a clear mandate: onboard 50 enterprise shippers in 6 months. The problem? Their existing shipper portal was a 4-year-old PHP monolith that crashed under load, had no mobile support, and took 8 seconds to load a shipment list.

Three enterprise prospects had already walked away citing the portal experience. The founding team estimated they were losing $2M in ARR to a bad UI. They needed a full rebuild — and they needed it before their next sales cycle.

Client Profile
CompanyTrellis Logistics
StageSeries A ($4M raised)
IndustryB2B Freight & Logistics
Team Size22 employees
EngagementFull-stack product build
Duration9 weeks
How We Did It

The process

01 Week 1–2

Discovery & Architecture

We embedded with the Trellis product team for two weeks of intensive discovery — mapping existing workflows, interviewing 12 shipper users, and auditing the legacy codebase. The output was a clear technical spec and a serverless AWS architecture designed to handle 10× the current load.

  • Technical architecture doc
  • User journey maps
  • API contract definitions
  • Risk register
02 Week 3–6

Core Platform Build

We built the shipper portal in Next.js with a serverless backend on AWS Lambda and API Gateway. Real-time shipment tracking was powered by WebSocket connections, while a PostgreSQL database on RDS handled transactional data. Every component was built with testing from day one.

  • Next.js portal (30+ screens)
  • Serverless API layer
  • Real-time tracking engine
  • Auth & RBAC system
03 Week 7–8

Integrations & Data Migration

The hardest part: migrating 4 years of shipment history from a fragile legacy system without downtime. We built a custom ETL pipeline, ran parallel systems for two weeks, and executed a zero-downtime cutover on a Sunday night. All 14 third-party carrier integrations were re-implemented with proper error handling.

  • Zero-downtime data migration
  • 14 carrier API integrations
  • ETL pipeline
  • Rollback plan
04 Week 9

QA, Hardening & Launch

A full week of load testing, security review, and user acceptance testing with 20 beta shippers. We resolved 47 issues before launch — none of them critical. The portal went live on schedule with a phased rollout starting with the top 10% of shippers by volume.

  • Load test report (10k concurrent)
  • Security audit sign-off
  • UAT with 20 shippers
  • Phased launch playbook
The Numbers

Results at a glance

11 wks

Faster time-to-market

vs. original 20-week estimate

9 wks

Actual delivery

from kickoff to production

99.98%

Uptime at launch

across first 90 days

Faster page loads

vs. legacy portal

Technology

Stack used

Frontend

  • Next.js 14
  • TypeScript
  • Tailwind CSS
  • React Query

Backend

  • AWS Lambda
  • API Gateway
  • Node.js
  • PostgreSQL (RDS)

Real-time

  • WebSockets
  • AWS IoT Core
  • Redis
  • SQS

DevOps

  • GitHub Actions
  • AWS CDK
  • CloudWatch
  • Datadog
Client Testimonial
“KeeMinds didn't just build what we asked for — they challenged our assumptions, caught three architectural mistakes before we made them, and delivered a portal that our enterprise prospects actually compliment during demos.”

Marcus Webb

Co-founder & CTO, Trellis Logistics

Build Something Like This

Ready to ship faster?

Whether you need a full rebuild or a greenfield product, we deliver production-grade software on aggressive timelines — without cutting corners.