Developer Tool

Import Preflight

Audit your CSV data, get a Clean Data Score, and auto-fix issues before import. 50+ detectors with semantic understanding of Shopify product structure.

Role Solo Creator
Stack Next.js 16, Supabase, Zustand
Detectors 50+ (12 core + 30+ Shopify)
Status In Development

What is Import Preflight?

Import Preflight is a SaaS platform that audits CSV data and generates intelligent fix recommendations before import operations. Upload a Shopify product CSV and the engine runs 50+ specialized detectors in parallel, returning a severity-based verdict (Safe / Needs Attention / Do Not Import) with a Clean Data Score from 0-100.

But it goes beyond detection. The auto-fix engine generates ~40 types of algorithmic fixes — no LLM required — that users can approve individually before applying. The platform builds a semantic understanding of your product data, inferring the product/variant/image hierarchy and detecting context-aware issues like orphaned variants and blank cell overwrites.

Silent failures, expensive consequences

Shopify's CSV import process is forgiving in the worst way — it accepts malformed data without clear warnings, leading to duplicate products, broken variants, missing images, and pricing errors that can take hours to untangle.

For agencies managing multiple client stores, a bad import can mean lost revenue, broken storefronts, and eroded client trust. The tooling gap became especially acute when the leading validation app was removed from the Shopify App Store, leaving merchants with no safety net.

50+
Parallel Detectors
~40
Auto-Fix Types
<5s
Typical Scan Time

Migration pain at scale

When migrating to a new platform like Shopify, merchants often have hundreds to thousands of products and variants to import. CSV files are the standard strategy because there are simply too many to enter manually. But according to users, these imports can take hours — and then randomly error out with no clear explanation of what went wrong.

We built Import Preflight to scan those files in 5-10 seconds and tell you upfront whether it would error out or have potential issues. It's designed as a B2B tool for agencies that are hired to handle these migrations and for individual merchants who prefer CSV importing and want to reduce risk with their product data.

"Hours of import time wiped out by one bad row. We wanted to catch that in seconds, not after the fact."

The stack

The platform runs on Next.js 16 with Supabase for auth, storage, and persistence. The scan engine parses CSV files, auto-detects file type (Shopify product vs. generic), and runs all 50+ detectors in parallel. State management uses Zustand, and large datasets are handled with TanStack Virtual for smooth scrolling through thousands of rows. The architecture is API-first — everything available in the UI is also available via REST endpoints for pipeline integration.

Framework
Next.js 16 (App Router)
Database & Auth
Supabase (PostgreSQL)
State
Zustand
Analytics
PostHog
Virtualization
TanStack React Virtual
Language
TypeScript (end-to-end)

The detection engine

01

Semantic Document Layer

Builds an internal product/variant/image graph from CSV rows. Understands hierarchy, infers CREATE vs UPDATE intent, and classifies row types — enabling context-aware detection no flat-file scanner can match.

02

50+ Parallel Detectors

12 core detectors (duplicates, whitespace, encoding, nulls) plus 30+ Shopify-specific detectors (orphaned variants, handle conflicts, option consistency, image URL validation, field length limits). All run in parallel.

03

Algorithmic Auto-Fix

~40 fix types — all heuristic, no LLM. Whitespace trimming, null string conversion, case standardization, date formatting, duplicate removal. Every fix shows before/after diffs for user approval.

04

Clean Data Score

A 0-100 score based on severity-weighted findings. Gives merchants an instant read on import risk — Safe, Needs Attention, or Do Not Import — with the hero finding highlighted.

What's next

The core scan engine, auto-fix pipeline, and rule engine are built and running. The landing page features a scroll-driven interactive demo showing the full scan flow without requiring sign-up. Current focus is on the paid tier, API documentation, and pipeline integration for agencies running bulk imports.

The architecture is built to expand beyond Shopify — the detector framework, rule engine, and scoring system are file-type-agnostic by design. New detector packs can be added for any CSV-based import system.

Back to all projects