Back to Case Study

AI Workflow Design

Status Draft v1.0
Owner Product, in partnership with ML Engineering
Series
Ad Copilot Portfolio Doc

This doc is the technical-product bridge between “we use AI” and “here’s exactly how the AI behaves, when it’s trusted, and what it costs.” It’s the document I’d expect to defend line-by-line with an ML engineer in a design review.

Shared pipeline for creative generation and performance recommendation:

Input → Model Call(s) → Confidence Scoring → Lane Assignment → Human Review (if required) → Action → Feedback Loop

1. Purpose

Two AI-driven capabilities exist: creative generation and performance recommendation. Both share the pipeline above. This doc defines each stage, the model limitations we design around, and the cost-control mechanisms baked into the pipeline rather than bolted on after.

2. Model selection & routing

Principle: match model cost/capability to the risk and reversibility of the output — not “use the best model everywhere.”

TaskModel tierRationale
First-pass creative generation (3–8 candidates)Lower-cost, fasterHigh volume, low individual stakes — filtered before a human sees them
Refinement of top-N candidatesHigher-capabilityLow volume; higher quality bar since these reach the reviewer
Drift detection & recommendation reasoningHigher-capability, strict output schemaNeeds defensible rationale over noisy time-series — what a marketer will actually judge
Brand-safety wordlist screeningDeterministic rules (not a model)Compliance-critical checks must not depend on probabilistic behavior

Known model limitations we design around

3. Prompt design principles

4. Confidence scoring methodology

Confidence is not the model’s raw self-reported score. It’s a composite:

Confidence = f(model self-score, historical acceptance rate, data volume, signal strength)

This composite feeds lane assignment (§5) and is recalibrated periodically against outcomes (§7).

5. Review lane logic

LaneCriteriaExample
Auto-applyHigh confidence AND blast radius below cap AND trust tier ≥ established5% budget nudge on a campaign with 60 days of stable accepted history
SuggestMedium confidence, OR blast radius above cap, OR newer accountMost v1 traffic — Suggest-only by design
EscalateLow confidence, OR insufficient data, OR compliance-sensitive actionRecommendation with <100 conversions of underlying data

Lane thresholds are configurable per account (US-11) but bounded by hard platform-level ceilings (e.g. no account can enable Auto-apply above a fixed % of daily budget). That ceiling is not exposed as configurable.

6. Cost controls

7. Feedback loop & evaluation

8. Deliberately out of scope