How it works
How Weldforge works.
You describe what to connect; we build it, deploy it, run it, and watch it — for a flat fee. Here's the whole thing end to end, plus pricing, security, and answers to the questions we hear most.
- 1
Describe the outcome
Tell us what you want connected in plain language — by chat, voice, or a short form. No field-by-field spec, no connector or auth decisions on your side.
- 2
AI drafts the mapping
The mapping wizard auto-drafts the field-by-field mapping, every field typed and previewed on real sample data, with plain-English rules and validation. You review and approve before anything is built.
- 3
We build & deploy
Senior architects plus AI build the integration and deploy it in our cloud — including any connector we need to build on the fly for your systems.
- 4
We run & you watch
We host, run, and monitor it. You watch runs, errors, throughput, and your live cost projection on a dashboard. If something breaks, it is our pager, not yours.
Frequently asked questions
What can Weldforge connect?
Any source to any target — SaaS apps, databases, file feeds, and in-house APIs. There is no fixed catalog you have to stay inside of; if a connector does not exist yet, we build it on demand as part of your build.
How is it priced?
A flat monthly fee per tier — Spark ($499/mo), Forge ($1,499/mo), or Foundry ($4,999/mo) — with annual discounts. No per-task or per-operation pricing, so cost does not spike when volume grows.
What happens if I exceed my plan?
You are alerted at 70%, 85%, and 100% of your usage, and no invoice ever exceeds your subscription without your one-click approval. No surprise overage bills.
How long does a build take?
Most integrations go live in one to three weeks, because the mapping is AI-drafted and reviewed with you before the build starts.
Is two-way sync supported?
Yes. Add the reverse as a second flow. The mapping wizard supports any direction, per-field rules, defaults, collections, nested objects, and a record-level sync filter.
How is my data kept secure?
Per-tenant data and credential isolation, encryption in transit and at rest, and an egress allowlist per integration. See the security page for the full posture; SOC 2 is on the roadmap.
Do I need engineers to use it?
No. The whole point is that you describe the outcome and we handle the build, the hosting, the running, and the monitoring. There is no tool for your team to learn or maintain.
Related: product overview, integrations catalog, security, pricing.
Reference docs — coming with V1
Detailed reference ships alongside the product. Until then, our team answers product and setup questions directly, usually within one business day.
Connector reference
Per-connector supported objects, auth, sync modes, and limits — published as each connector ships.
Dashboards & monitoring
Reading runs, errors, latency, throughput, and your live cost projection.
API & webhooks
Programmatic access and event hooks for teams that want them.