Vermont Systems · Internal

MainTrac Docs

Research, architecture, and setup documentation for the MainTrac standalone product initiative.

RecTrac 3.1.11.10.00 Cohort 1 Decomp Candidate Last updated May 2026
Initiative
MainTrac Extraction
Decomp Status
Cohort 1 Confirmed
Current Customers
~80 organizations
Tech Stack Decision
Complete
Dev Environment
Mac Mini M4
Documents
📋

Product Overview

Comprehensive product document covering all functional modules, data model, technical architecture, RecTrac coupling analysis, decomposition assessment, known gaps, and version history. Written for leadership and product group.

⚙️

Tech Stack Analysis

Deep-dive into building MainTrac from scratch as a modern web and mobile app. Covers framework choices, mobile UX, database design, auth, real-time, scale, multi-tenancy, deployment tiers, and migration path from Progress/OpenEdge.

Role Guides
🏛️

Leadership

Scope approval, budget authorization, deployment model policy, customer migration timeline, cross-brand strategy, and App Store distribution decisions.

📦

Product

MVP definition, mobile-first sequencing, beta customer selection, gap prioritization, contract processing, API scope, naming, and migration order.

⚙️

Engineering

Auth provider, cloud provider, CI/CD tooling, connection pooling, WatermelonDB spike, ETL strategy, feature flags, and TypeScript standards.

🧪

QA

Test strategy, multi-tenant isolation, mobile device testing, migration validation, performance targets, security testing, accessibility, and test environments.

🏗️

Architect

DB naming, RLS session strategy, OpenAPI standards, monorepo boundaries, offline sync conflict resolution, background jobs, observability, JWT, caching, and DR targets.

Context

What is this initiative?

MainTrac is Vermont Systems' maintenance tracking module, currently embedded in RecTrac 3.1. In May 2026, it was confirmed as a Cohort 1 decomposition candidate — a leaf node with no shared engine dependencies and no unified transaction table entries. The goal is to extract it as a standalone product and open it to NextRec and RecDesk customers.

Recommended stack (summary)

  • Web UI: Next.js 15 + shadcn/ui + Tailwind
  • Mobile: Expo (React Native) + WatermelonDB offline
  • API: Hono on Bun, REST + OpenAPI
  • Database: PostgreSQL 16 + Drizzle ORM + RLS
  • Auth: Clerk (org-level multi-tenancy)
  • Hosting: AWS ECS Fargate + RDS + RDS Proxy
  • Monorepo: Turborepo + pnpm

Key decisions made

  • Native mobile app (Expo) over PWA for field workers
  • Shared SaaS deployment (RLS) with optional dedicated tier
  • Modular monolith to start — not microservices
  • PostgreSQL RLS for multi-tenancy across ~80 customers
  • RDS Proxy required for connection pooling at this scale
  • Ship mobile app before web redesign (highest customer pain)

Open questions

  • Which customers need dedicated tier (CNIC / data residency)?
  • Clerk vs Auth0 — check existing VSI vendor contracts
  • Railway for staging vs AWS from day one?
  • Historical data volume per customer (profile top 10 first)
  • iOS App Store vs enterprise distribution for field workers