External systems that connect, retry, and stay in sync — without hand-built plumbing.

She sets up what makes an outbound integration run and stay resilient — the connection to the external system, its operations, stored credentials, and the resilience settings that keep transient failures from breaking a flow. She wires those into your flows so the right call fires on the right event. The live traffic and records stay in your hands; Phyllis builds the pipes, not the messages.

Always first. Before building anything she takes stock of your existing connections, credentials and installed vendor connectors. If a ready-made connector already covers the system you're integrating — Slack, Jira, a cloud provider — she'll prefer it over hand-building, because it already handles the hard parts. She only creates new plumbing when nothing suitable exists.

Carefully, and never by guessing. She keeps credentials in centralized profiles rather than inline, so the same call works across dev, test and prod without duplicating secrets. If a build needs a secret or token she doesn't have, she stops and asks — she never fabricates credentials or assumes an authentication type.

Yes — it's one of the first things she confirms. Some integration capabilities depend on licensing that can't be detected reliably, so she checks what you're entitled to before building and matches the approach to it. Either way you get a working integration that fits what you're licensed to run — no surprises later.

Correct. Phyllis reviews your instance, cites every finding, proposes a plan, and only builds on your explicit approval. Every change is staged for review and is reversible, and she never completes or activates anything on your behalf. You decide what goes live.