# FAQ — bild

> Fair questions about agent workflows, checkpoints, data and how an engagement starts.

Markdown rendering of https://bild.systems/faq — the HTML version is canonical. Contact: humans@bild.systems

# Fair questions.

The ones we get asked most, answered the way we'd answer them in the room.

## Will this replace my team?

The agent drafts. Your team approves.

Workflows return hours to the people you hired to think. Nothing we build removes a decision from a person — every point where judgment matters gets a named checkpoint, and that's method, not marketing.

## Our processes are too messy for automation.

That's the starting point, not a blocker.

The audit exists precisely for this. Two weeks inside your operations finds the work worth automating and the work that isn't — and if nothing is worth building, the audit says so.

## We tried automation before and it broke.

Exceptions surface at a checkpoint instead of failing silently.

The usual failure is silence: something changes upstream and nobody notices until the numbers are wrong. Our workflows flag the exception, stop at a checkpoint, and wait for a person. Every run leaves a record.

## Is our data safe?

Your stack, your guardrails, your approvals.

Workflows are built on the tools you already run, inside the security boundaries you already trust. Nothing leaves the room it already lives in, and we resell no platform — so no vendor margin steers the architecture.

## How will we know it worked?

We agree the number before we start.

Hours returned, exceptions caught, volume handled — the measure goes in the proposal, not the retrospective. You'll know whether it worked without taking our word for it.

## How do we start?

Tell us what's stuck.

Describe the work that keeps getting in the way — the spreadsheet, the routing, the chasing. The audit takes two weeks, and it opens every engagement.

A question we haven't answered? Ask a human.