Turbo AI PM
Kits / workflow

Customer-facing chatbot / support

Talks to customers directly, often with no human in the loop. Work top to bottom, or jump to the decision in front of you — each one links to the metrics, tool, and playbook that answer it.

The riskDeflecting customers instead of resolving their problem.
Used in product scenariosCustomer Support
Take it to your PRD, Notion, or launch review.

The chain

What has to be true at each layer for this product to create value. The hypotheses between layers are yours to hold — and to check after launch.

Model / SystemIs the AI behaving correctly?
R-05 Groundedness / Faithfulness↔ R-07S-01 Hallucination Rate↔ S-03P-01 Latency (p50 / p95 / p99)↔ C-01
If answers are correct and fast, customers get resolved without an agent.
Product / UserIs the user accomplishing the task?
B-03 Deflection / Containment Rate↔ B-05B-08 Time to Resolution Delta (TTR Δ)↔ B-05B-05 CSAT / Thumbs-up Rate↔ B-03
If resolution is real rather than deflection, cost per ticket falls without hurting satisfaction.
BusinessIs the product creating meaningful value?
Cost per ticket · repeat-contact rate · churn

↔ marks each metric's shadow metric — the one to watch alongside it so the number can't improve while the product gets worse.

01

Before you build

Should this be AI at all — and which capability fits?

Before scoping a model, check that the problem has a pattern to learn, data you can reach at decision time, and a cost of being wrong you can live with. If a rule would get most of the value, the rule is the better product.

For Customer-facing chatbotThe test for a support bot: are most tickets repetitive questions with documented answers? If every ticket is bespoke, you are automating the easy 20% and keeping all the hard work.
Check whether this needs AI →
Evidence, business implication, deeper reading
Evidence you need: Examples of the decision being made today, where the data lives and whether it is available at decision time, and an honest estimate of what a wrong answer costs.
Business implication: Choosing AI for a problem rules could solve buys you cost, latency, and a maintenance burden with no extra value.

Check whether the unit economics can work

Estimate cost per completed task against the value of that task before building — margins are set by architecture, not tuning.

MetricsC-02 Cost per Resolved Task↔ B-02C-01 Cost per Inference↔ P-01
Estimate ROI & unit economics →
Evidence, business implication, deeper reading
Evidence you need: Volume estimate, token estimate per task, model pricing, and the human cost (or revenue) per task.
Business implication: A feature that works but destroys gross margin is a failed product.

Make the case to build it

Ask for one specific decision by a specific date: the problem in their units, why AI and not rules, the expected value with its assumptions visible, the risk said out loud, and the checkpoint you will report back at.

MetricsB-03 Deflection / Containment Rate↔ B-05
Build the business case →Make the case & defend the work →
Evidence, business implication, deeper reading
Evidence you need: The baseline (what today costs and achieves), a value range with the two assumptions that move it most, cost per completed task vs. the human cost, the rules-vs-model comparison, and the top failure modes with their mitigations.
Business implication: A named checkpoint turns a bet into a staged decision — much easier to approve than an open-ended one. For this product: Cost per ticket · repeat-contact rate · churn.

Define what the model may do, must never do, and needs to know

Write the operating envelope before anyone writes a prompt: context the model sees, unforgivables, and what happens when it fails.

For Customer-facing chatbotWrite the unforgivables: no refunds promised, no account changes without verification.
Write the AI PRD →
Evidence, business implication, deeper reading
Evidence you need: Agreed context spec, a written list of unforgivables, and a fallback for each failure mode.
Business implication: Scope decided here drives both cost and risk for the life of the feature.
02

Evaluate

Choose what to measure

Pick one metric per layer and a shadow metric for each, so no number can improve while the product gets worse.

MetricsB-03 Deflection / Containment Rate↔ B-05B-08 Time to Resolution Delta (TTR Δ)↔ B-05B-05 CSAT / Thumbs-up Rate↔ B-03S-03 Refusal Rate↔ B-02S-02 Toxicity / Harmful Content Rate↔ S-03P-01 Latency (p50 / p95 / p99)↔ C-01
Calculate any of 38 metrics →
Evidence, business implication, deeper reading
Evidence you need: The starter metrics below, each paired with the metric that catches it being gamed.
Business implication: The layer-3 metric is the one leadership will ask about — name it now. For this product: Cost per ticket · repeat-contact rate · churn.

Establish a baseline worth beating

A number means nothing without a comparison. The simple heuristic and the incumbent are the two baselines that get skipped.

For Customer-facing chatbotYour existing help-center search or macro suggester, plus human agents' resolution rate and time.
MetricsR-05 Groundedness / Faithfulness↔ R-07B-03 Deflection / Containment Rate↔ B-05
Evidence, business implication, deeper reading
Evidence you need: Zero-rule, heuristic, and incumbent scores on the same eval set, measured the same way.
Business implication: If a heuristic gets 80% of the value, the model has to justify its extra cost.

Design an evaluation you can trust

Representative data, a scoring method that matches the task, and a holdout nobody tunes against.

MetricsR-05 Groundedness / Faithfulness↔ R-07S-01 Hallucination Rate↔ S-03P-01 Latency (p50 / p95 / p99)↔ C-01
Evidence, business implication, deeper reading
Evidence you need: A golden set with an owner and a size target, slices defined up front, and a judge calibrated against humans.
Business implication: A weak eval means you ship blind and cannot explain failures afterward.
03

Ship

Decide whether it is ready to ship

The bar depends on blast radius and reversibility, not on the accuracy number alone.

For Customer-facing chatbotCustomer-facing at scale is amber at best. Human escalation must always be one step away.
MetricsR-05 Groundedness / Faithfulness↔ R-07S-01 Hallucination Rate↔ S-03
Decide go / no-go →Make the case & defend the work →
Evidence, business implication, deeper reading
Evidence you need: Your quadrant, the threshold for it, the eval result against the baseline, a working fallback, and the numbers you promised at the funding checkpoint.
Business implication: A documented bar turns post-launch blame into post-launch learning.

Design what users see when it is wrong

Every AI flow needs a designed failure state: specific copy, a way to recover, and a human path.

For Customer-facing chatbot"I couldn't find that — want me to connect you to someone?" with the conversation handed over intact.
Design for non-determinism →
Evidence, business implication, deeper reading
Evidence you need: Fallback copy for timeouts, low confidence, out-of-scope requests, and refusals.
Business implication: Users forgive visible, recoverable failures; they abandon silent ones.
04

In production

Set up monitoring that separates "the model changed" from "the world changed"

Alert on breakage, review drift daily, review quality weekly, review business impact monthly.

MetricsR-05 Groundedness / Faithfulness↔ R-07S-01 Hallucination Rate↔ S-03P-01 Latency (p50 / p95 / p99)↔ C-01
Evidence, business implication, deeper reading
Evidence you need: Error and latency alerts, input-mix and confidence drift, slice quality, and the layer-3 metric on a monthly cadence.
Business implication: Degradation you catch in a week costs far less than one you find in a quarterly review.

Turn production signal into a better next version

Design what each interaction captures and where it goes — implicit, explicit, and outcome signal feeding the eval set.

For Customer-facing chatbotEscalated conversations are labeled examples of what the bot cannot do yet.
MetricsB-03 Deflection / Containment Rate↔ B-05B-08 Time to Resolution Delta (TTR Δ)↔ B-05
Evidence, business implication, deeper reading
Evidence you need: A named owner for the signal pipeline and a path from user action to golden set.
Business implication: The flywheel is the only part of an AI feature that compounds over time.
!

Something looks wrong

Start from what you're seeing. Each symptom points to its most likely cause and the metrics that confirm it.

Deflection up, repeat contacts up
Customers are being deflected, not resolved — they give up and come back.
CheckB-03 Deflection / Containment Rate↔ B-05B-05 CSAT / Thumbs-up Rate↔ B-03Investigate →
Average looks fine, enterprise customers complain
A slice the average hides.
CheckB-05 CSAT / Thumbs-up Rate↔ B-03Investigate →
The bot refuses too often
Over-tuned safety settings blocking legitimate requests.
CheckS-03 Refusal Rate↔ B-02B-02 Task Completion Rate↔ B-06

Other kits

RAG / knowledge assistant Agents & multi-step pipelines Copilot / drafting tool Classifier / triage & moderation Coding assistant / IDE copilot