Orientation

New here? This site is the internal explainer for PulseAd and its product. This page is the reading path; follow it in order and you’ll have a working mental model in about fifteen minutes.

Anchor on these two facts first

Almost every confusion traces back to one of these:

1. The names overlap. PulseAd (company/platform) ≠ Pulsy (the product) ≠ PULSEAD/PULSY (Snowflake schemas) ≠ Pulson/PSM (older/legacy chat names) ≠ AOP (the internal operator tool). The name map untangles all of it — read it before anything else.
2. Pulsy reads freely but acts only with approval. A chat reply that says it changed something has only proposed a change; a human approves before anything reaches Amazon. The whole gate model is in Capabilities.

The reading path

1. What PulseAd is — the company, the name map, the product surfaces, and the one behavioral rule. Start here.
2. Capabilities — what Pulsy can read vs. change, and the three gates every write passes (including the pilot-only “live” brand).
3. Data model — what Pulsy knows: the six analysts over Snowflake, and how current the data is.
4. Data & coverage — which channels are covered, Amazon vs. the manual off-Amazon upload.
5. Creative — the image and video generation tools.
6. Reference + the glossary — the terms the rest of the docs assume.

How these docs work (so you can trust them)

• They’re derived from the live source repos, pinned to the commit they were generated from, and re-syncable. The full procedure and writing rules are in AGENTS.md.
• They document the truth candidly — including legacy surfaces and known gaps — because this is internal. If you find something asserted that the code doesn’t support, that’s a bug; flag it.
• Found a gap the docs can’t answer? Open questions and “verify with ops” items live in _internal/needs-johnny.md — add to it rather than guessing.