Reference Architecture · Conceptual

Automated experiment health checks, without an application

A daily agent that reads A/B test configs, computes anytime-valid statistics, narrates them for a non-technical audience, and delivers the result — running entirely as a warehouse transformation DAG plus one delivery function. Hover any component or wire for what it does and why. Click to pin; use the flow buttons to trace one path.

no orchestration framework · no server sequential anytime-valid inference prompts stored as data, not code conceptual · anonymized
Trace a flow hover = inspect · click = pin

Why it's built this way

The decisions that matter, and the reasoning behind each. Several invert what you'd reach for by default.

AUTOMATION FORCES BETTER STATISTICS

A human checking an experiment weekly gets away with fixed-sample tests. A system that re-evaluates every metric every day is continuous peeking — Type-I error inflates far above nominal, silently, with the authority of automation. Anytime-valid confidence sequences are what make the concept legitimate at all.

THE MODEL NARRATES; IT NEVER INFERS

Every statistical decision — significance, evidence tier, gates — is computed deterministically in SQL before the prompt is assembled. The model receives a JSON object where the hard calls are already made and translates it into prose. That's precisely why an LLM is appropriate here.

BAN RAW STATISTICS FROM THE OUTPUT

No p-values, no power, no confidence intervals. Instead: uplift, the underlying rates, sample sizes, and plain-language reliability. The audience is PMs and leadership, where a p-value isn't information — it's a Rorschach test. Precision in output is not the same as usefulness.

CLASSIFY EVIDENCE BEFORE WRITING

Each metric is sorted into strong / moderate / weak / no-effect / harm before any prose is generated, and only the top tier licenses a confident rollout call. "Not significant with adequate power" is framed as informative — evidence of no meaningful effect — not lumped in with "inconclusive."

PROMPT-AS-DATA CUTS BOTH WAYS

Templates in a warehouse table mean editing a row deploys a new prompt — seconds instead of hours, and a non-engineer can improve the output. It also puts the most behavior-defining artifact in the system outside version control, with no review or rollback. Only do this if you also build the eval loop.

SOFT-FAIL GATES GET READ

The sample-ratio-mismatch check flags loudly and shows the metrics anyway rather than withholding them. A gate that silently blocks results teaches people to route around the tool; one that caveats loudly gets taken seriously.

FAIL CLOSED AT THE DELIVERY BOUNDARY

A publish gate admits only complete results stamped with the current date, and delivery inner-joins it. A broken run therefore ships nothing. Silence is an acceptable failure mode; confident staleness is not.

MEASURE AGREEMENT, NOT JUST QUALITY

The eval loop regenerates each check ~10× and checks whether the recommendation and direction hold across runs. A prompt that flips its recommendation on identical data is disqualified regardless of how well any single run scores — instability that a mean quality score misses entirely.

SAY WHAT THE NUMBER MEANS

The estimand is declared explicitly — effect among users who actually saw the change, not everyone assigned. "+6.5%" means different things under different estimands, and a reader can't tell which unless the system says so.

ENUMERATE YOUR NON-GOALS

The methodology spec names what it deliberately doesn't do — variance reduction, FDR, multi-arm correction, intent-to-treat — each with its reason and mitigation. A spec that admits its gaps is more trustworthy than one implying completeness, and it stops the next engineer assuming guarantees that aren't there.

MIGRATE INFERENCE IN SHADOW MODE

Replacing the engine ran new-alongside-old invisibly, then A/A validation to confirm the false-positive rate matched nominal, then a cutover that froze methodology for in-flight tests. Changing how significance is computed mid-experiment invalidates the experiment.

CONFIG BELONGS WHERE THE HUMANS ARE

Experiment definitions live in the docs tool analysts and PMs already use, replicated by an off-the-shelf connector. Zero admin UI to build, zero bespoke sync code, and PMs can review config without warehouse access.