CSM Health & Renewals Agent

The approach.

Capture the real org, write the PRD, let AESOP run the build with a platform choice — then prove it on synthetic data. The org is real; only the data is fake — and that’s enough to stand up a working demo, fast.

01 · The approach
1

Discover

01-discover-brief

Capture the org — the real before-state (4 hrs/wk rebuilding health by hand, risk surfacing late), the real stack, the real constraints. End in open questions, not assumptions.

2

Build PRD

working agreement · design spec · build PRD

The spec set: a governance contract, the architecture and prompt-vs-code line, and the build-ready PRD. Rules locked before code.

3

AESOP build

build-artifacts

The PRD feeds AESOP, which runs the build and makes a platform choice: FastAPI on Fly (backend), Next.js on Vercel (frontend), Supabase for data + auth, and a provider-flexible LLM (DeepSeek / Anthropic).

4

Synthetic data, real org

data/ · seed 42

A 12-account book exercises every rule and both draft framings. Synthetic data removes the blockers — no real customer data leaves the governed environment, no approval to wait on — so it’s fast enough to stand up a demo that’s still grounded in a real, captured org.

02 · The prompt-vs-code line

The decision that governs every stage: what lives in deterministic code and what the model is allowed to do.

DecisionWhereWhy
Risk scoring & thresholdsCodeReproducible, auditable, cheap
Escalation routingCodeDeterministic — never skipped
Grounding checkCode + LLM judgeDeterministic trace + semantic check
Health narrativePromptLanguage synthesis is the model’s job
Renewal draftPromptTone + framing, human-reviewed
03 · The principles that govern every stage
Deterministic-first

The risk math is code, never the LLM. A flag means the same thing every run, and every claim traces to a signal or a quoted sentence.

Grounded

Nothing asserted without a receipt. Missing usage is reported as a gap, never fabricated into a trend.

Human-in-the-loop

Everything customer-facing stays draft. The CSM confirms every flag and edits every draft — the agent never sends.

Honest eval

Self-scored weekly, and not a rigged 100% — escalation accuracy is pinned at 85.7%, because one judgment-boundary case is the point of the eval.