FAQ

Straight answers, before you ask

The questions evaluations actually start with. Anything missing? Ask directly. Answers come from engineering, not a script.

The product

What is SpecLatch, in one sentence?
SpecLatch is one place for the technical data of the products a company makes or distributes: specifications live once as typed, unit-aware fields, with formulas and confidentiality built in, and that product data then feeds your spec sheets and catalogs, a live website, an AI assistant, and the distributors who sell your products.
Who is it for?
Manufacturers and distributors of specification-driven products: companies whose buyers ask for a spec sheet before they ask for a price. It's built to be run by the people who own the data (engineers, product managers, marketing) rather than by developers. Tables, fields, formulas, documents, and publishing rules are all configured in the interface.
Is SpecLatch a PIM?
That's the nearest shelf label, but SpecLatch is built for engineering data in ways general PIM tools aren't: fields know their units and convert between them, calculations are formulas checked for units before they save, every field carries a confidentiality tier the database applies everywhere, and documents are generated from live data.

Data & files

What can I import?
CSV and Excel files. Columns map to fields automatically by header name, with optional AI suggestions for the rest on Growth and up, and source units convert on the way in, so a kg column lands correctly in a field stored in lb. A dry run shows every change before a single row is written. Once data is in, the grid also accepts paste straight from a spreadsheet.
What file types can I attach to records?
Images, PDF, Office documents, CSV/TXT, ZIP and CAD files. That covers supplier datasheets, drawings, CAD exports, photos, and certificates; other file types are refused at upload. PDFs and plain-text files also get their text extracted, so what's inside a document is searchable. Attachments carry their own visibility flag, so a confidential drawing stays internal even when it hangs off a public record.
Can I get my data out?
Always. One click exports the entire organization to XLSX, with every table, record, link, and document reference, and it works during a trial, after one lapses, whenever. Generated documents are standard PDFs. There are no proprietary formats anywhere in the exit path.
Which units does it handle?
Flow, pressure, power, torque, speed, temperature, dimensions, and more. A value is stored once in a primary unit and displayed in both imperial and metric: gpm alongside m³/h, psi alongside bar, HP alongside kW. Conversions are tested code, so there are no per-cell formulas for anyone to maintain.

Access & security

Who can see what?
Every field carries one of three confidentiality tiers (public, internal, or confidential), set once on the field definition. Every surface reads the same tier: the grid, generated PDFs, the public portal, and the API all show a given audience exactly the fields that audience is entitled to. An approval workflow sits between draft and published. Drafts and records under review stay off the portal and the public API until they're published.
Where does my data live?
In a managed PostgreSQL database, with files in object storage (Cloudflare R2), on infrastructure with automated backups. Every table is scoped by organization ID, so your organization's data is isolated from every other tenant's. The security page describes the enforcement mechanics in more detail.
Do AI features train on my data?
No. The AI suggestions (import column mapping, plain-English filters) are opt-in: they run against your own Anthropic or OpenAI key, added in settings. Without a key, they are off. SpecLatch doesn't train models on your data.
Can an AI assistant answer questions from my catalog?
Yes, from the Growth plan. SpecLatch is an MCP server, so Claude, Cursor, Copilot, or any MCP client can connect with one of your API keys and answer from your published records. The key's tier caps what the assistant can read, so a public key sees public fields only. Access is read-only.

Getting started & pricing

How do I get started?
Sign up and create a workspace. There is no invite, no sales call, and no credit card. Every new workspace starts with a 30-day trial with Growth-plan features, so the fastest evaluation is importing one of your own product families and seeing your own data in it. If you'd rather talk to a person first, the contact page reaches the team directly.
What does it cost?
Three plans, published in full on the pricing page: Starter at $49, Growth at $149, and Professional at $399 per month. Annual billing is ten months' price, so two months are free. Enterprise is priced per conversation. Viewers are unlimited and free on every plan, and they can open the generated spec sheets and catalogs; only roles that can change data take a seat.
Is there a free trial?
Yes. Every new workspace starts with a 30-day trial with Growth-plan features, no credit card required. Unused trial days carry over when you subscribe: pick a plan on day 10 and the remaining 20 days come with you before billing starts. If a trial lapses, the workspace becomes read-only. Nothing is deleted, and export keeps working.
Can I migrate from spreadsheets or another tool?
That's the expected starting point. Export what you have to CSV or Excel, import it with automatic column mapping and unit conversion, and check the dry run before committing. Most teams start with a single product family rather than migrating everything at once. The first one proves the structure, and the rest follow it.

Rather see it than read about it?

Import one of your own product families. The trial is the demo, on your data.

Start free trial