Dispatch Compliance Engine

The single source of truth for field service dispatch teams.
Krist Watson, Founder  ·  dispatch-compliance-engine.vercel.app
Nebius for Startups Application — May 2026

The Problem

Dispatch coordinators burn 4-6 hours per day reconciling broken systems.

Today's Reality

Dispatch teams juggle Smartsheet, Field Nation, Service Channel, customer portals, and email — manually copying data between them.

The Cost

Missed deadlines, billing clawbacks, double-booked techs, compliance failures — single missed check-in costs $200-$2,000.

Why Nothing Works

ServiceTitan = $400/seat, built for HVAC. Field Nation = workforce only. Smartsheet = spreadsheet. None glue them together.

Target market: 12,000+ MSPs and IT service firms in North America handling Amazon, Whole Foods, retail, and enterprise on-site work.

The Product

One workspace. Real-time sync. RBAC enforced.

What DCE does today

  • Unified dispatch record — site, tech, time, attachments
  • Bidirectional Smartsheet sync (per-customer column mapping)
  • Manager / dispatcher / tech / accounts RBAC via Supabase RLS
  • Full-text search across dispatches, sites, techs
  • Reconciliation: cost, billing, margin, attachment audit

Stack

  • React + Vite frontend, Vercel hosted
  • Supabase Postgres + RLS + Edge Functions
  • Smartsheet API integration with field-level whitelist
  • Field Nation, Service Channel adapters in progress
● LIVE in production 426/426 tests · 17/17 e2e

Where We Are

Production-ready. Validating with first customers.

12Months building
426Tests passing
$0Outside capital
1Pilot customer

Q2 2026 — NOW

  • Pilot deployment with TechNet Partners (active)
  • Field-test reconciliation with real billing data
  • Onboard 3 paying MSPs at $300-$800/mo

Q3-Q4 2026 — AI Features

  • pgvector embeddings → dispatch similarity search
  • Hermes-4-70B → tech notes summarization, billing anomaly detection
  • Auto-classify customer requirements from intake docs

Why Nebius

The AI features need real inference at production scale.

FeatureVolumeModelMonthly Cost
Tech-note summarization (every dispatch close)~5K calls/moHermes-4-70B~$80
Billing anomaly detection (per invoice)~2K calls/moHermes-4-70B~$40
Dispatch similarity (pgvector + embeddings)~50K embeddings/moEmbedding endpoint~$30
Intake doc classification (per onboarding)~500 calls/moHermes-4-405B~$60
Production target~$210/mo

Why Nebius specifically: Hermes-4 family on Token Factory is 10x cheaper than OpenAI for native tool calling, native EU + US presence (compliance for Whole Foods customers), and you actually have inference economics that work for SaaS — not GPU rentals priced for training.

The Ask

$5,000 in credits → 18 months of AI feature dev + first 10 paying customers.

$5K

Nebius for Startups credits

What it unlocks

  • Build + ship 4 AI features without burning runway
  • Cover production inference for first 10 paying customers
  • Convert pilots into reference accounts before raising
  • De-risk the AI roadmap before customer-funded rollout

Why this works for Nebius

DCE will be a long-term Token Factory customer with predictable, growing inference spend tied directly to MSP customer count. Every paying MSP = $20-40/mo additional inference burn.

Year 1 target: 50 paying MSPs → ~$1,500/mo Nebius spend

Year 2 target: 250 MSPs → ~$8K/mo Nebius spend

Let's Talk

Thank you.

Solo founder, shipping product, AI-native by design.

Krist Watson — Founder, Dispatch Compliance Engine

dispatch-compliance-engine.vercel.app

github.com/KrisW360/dispatch-compliance-engine

Email: kwatson@technetpartners.com