CPMAI practice questions
PMI · CPMAI · 200 questions
Original practice questions for the PMI-CPMAI (Cognitive Project Management in AI) v7 exam, covering AI fundamentals (core AI/AGI/cognitive-computing concepts, the seven patterns of AI, and machine learning fundamentals), the CPMAI methodology (why AI projects differ from traditional software projects, and the six-phase lifecycle: Business Understanding, Data Understanding, Data Preparation, Model Development, Model Evaluation, and Operationalization, including Go/No-Go assessments at each phase), machine learning (classification and clustering, neural networks and deep learning, generative AI and large language models, and ML tools/platforms), data for AI (Big Data concepts, data governance, data pipelines, and data preparation/transformation), managing AI (model performance evaluation and production deployment), and trustworthy AI (ethics, privacy, security, transparency, and AI regulation) - framed inside Meridian Regional Airlines, a mid-size regional carrier running AI initiatives across predictive maintenance, crew and gate scheduling, customer-service chat, loyalty-program fraud detection, demand forecasting, computer-vision ramp safety, and a generative-AI knowledge assistant for maintenance and dispatch documentation.
This course contains the use of artificial intelligence.
About the CPMAI exam
- Time allowed
- 2 hours
- Questions
- 100
- Format
- Computer-based (Pearson VUE) or Online-Proctored Testing; 100 questions (90 scored, 10 unscored pretest), pass/fail scoring only
Exam details published by the vendor, checked 3 September 2026. Vendors change fees and formats without notice — confirm on the vendor's own page before you book.
Practice Quizzes
Test your knowledge with standard 20-question practice sets.
Quiz 1
Quiz 2
Quiz 3
Quiz 4
Quiz 5
Quiz 6
Quiz 7
Quiz 8
Quiz 9
Quiz 10
Browse by Domain
Study specific topics at your own pace.
AI Fundamentals · 32 questions
- Meridian Regional Airlines' new VP of Digital asks the project team to define 'artificial intelligence' for a board slide before funding is approved. Which definition best matches how CPMAI frames AI relative to human intelligence?
- Meridian's maintenance director hears vendors pitch both 'AGI' and 'narrow AI' for the predictive-maintenance program and asks the PM to clarify the difference before signing a contract. What should the PM say?
- Meridian's legal team asks whether the customer-service virtual assistant should be evaluated with the Turing Test before launch. What is the correct guidance on the Turing Test's role here?
- Meridian's dispatch team wants a system that can combine natural-language understanding, reasoning over structured flight data, and pattern recognition to help controllers make faster rerouting decisions during weather disruptions. Which AI concept best names this kind of system?
- During a steering-committee meeting, a finance executive says, 'AI will completely replace our maintenance planners within a year, and it never makes mistakes.' As the AI project manager, which response correctly debunks this common misconception?
- Meridian's dispatch operations still use a decision-support tool built in the early 2000s. The current PM proposes an AI refresh but a veteran manager warns, 'We tried something like this back in the 1980s and it went nowhere.' What historical AI concept explains that earlier failure and should inform today's risk plan?
- Meridian's ramp-safety team built a rule-based system in the 1990s that flags foreign-object debris only when it exactly matches one of 200 pre-coded object shapes, and it has never been updated to reason about partial matches or uncertainty. What kind of system is this, and what is its key limitation?
- Meridian's leadership asks why AI initiatives are suddenly a board priority now, when the airline has used data analytics for two decades without an 'AI' label. Which explanation is most consistent with CPMAI's framing of why AI has gained prominence in the current era?
- Meridian's CIO wants the AI roadmap folded into the airline's broader digital-transformation initiative rather than run as a side project. From a CPMAI standpoint, what is the strongest justification for connecting AI initiatives to the digital-transformation program?
- A junior analyst on the loyalty-fraud project says, 'Augmented intelligence and artificial intelligence are the same thing — the terms are interchangeable.' How should the PM correct this for the fraud-detection use case?
- Meridian's board asks the PM to distinguish, in one sentence, 'strong AI' from 'weak AI' as those terms are used in the exam content the PM studied. Which statement is accurate?
- Meridian's cargo-planning director wants an AI system that can guarantee, with zero error, the exact demand for every route six months in advance, and expects to cancel the project if it cannot hit that guarantee. What should the PM tell the director about this expectation before scoping the project?
- Meridian's HR department wants an AI system to make final termination decisions about ground-crew safety violations based solely on camera footage, with no human review. Which response correctly applies CPMAI's guidance on identifying unsuitable AI applications?
- Meridian's product team is scoping seven different AI ideas and asks the PM to map each to one of CPMAI's seven patterns of AI. Which pairing correctly matches a business need to its pattern?
- Meridian wants to deploy a virtual assistant that can hold a multi-turn conversation with a passenger about a delayed connecting flight, ask clarifying questions, and offer rebooking options. Which AI pattern is this, and what capability does it primarily require?
- Meridian is evaluating a vendor's speech-to-text system for pilot radio communications, a text-to-speech system for automated gate announcements, and a translation tool for European codeshare documents. Which statement correctly groups these three capabilities?
- Meridian's ramp-safety team deploys a computer-vision system that watches live camera feeds and alerts a supervisor whenever a ground-crew member enters a marked danger zone while an aircraft's engines are running. Which of the seven patterns of AI does this best represent?
- Meridian's ops team wants to automate two different things: (1) copying passenger rebooking data from an old reservation system into a new one whenever a flight is cancelled, and (2) predicting which flights are likely to be cancelled next week due to weather. A junior PM says both should be built as 'unattended bots.' Where is this wrong?
- Meridian wants a system that autonomously determines the best sequence of gate reassignments, crew swaps, and delay absorptions to minimize total passenger disruption during a weather event, adjusting its plan as new constraints appear. Which pattern best fits, and why?
- Meridian's customer-service AI program combines a conversational assistant for passenger questions, an anomaly-detection model for loyalty fraud, and a predictive model for cargo demand, all reporting into one shared operations dashboard. From a CPMAI standpoint, what should the PM keep in mind when integrating multiple AI patterns into one comprehensive application?
- Meridian's operations team wants an unattended bot to automatically rebook a passenger whenever their connecting flight is cancelled, without requiring a human to trigger it, but wants a human gate agent to manually approve any exception involving a passenger with a disability accommodation. How should the PM describe this design using RPA terminology?
- Meridian's new PM is drafting a one-slide explainer for the maintenance director and writes, 'AI is built on top of machine learning, which is its cornerstone.' A skeptical engineer disputes the ordering. Which statement is accurate?
- Meridian's fraud-detection team is choosing between a rule that says 'always flag a redemption over 50,000 miles' (deterministic) and a model that outputs 'this transaction has an 82% likelihood of being fraudulent' (probabilistic). What is the key project-management trade-off the PM should explain to stakeholders?
- Meridian's crew-scheduling optimization tool can solve small scheduling scenarios by trying every possible combination, but for a full week of flights, that approach would take longer than the scheduling window allows. A data scientist proposes a heuristic method instead. What should the PM understand about this trade-off?
- Meridian's demand-forecasting model outputs a specific number of passengers expected on a route next month (prediction), a data scientist explains why the model believes fuel-price trends are driving that number (inference), and a separate check confirms the model performs well on routes it has never seen historical data for (generalization). Why does the PM need to keep these three concepts distinct when evaluating the project?
- A vendor tells Meridian's PM, 'our algorithm and our model are the same thing, so once you approve the algorithm, the model never needs re-review.' What is wrong with this statement, and how should the PM correct it?
- Meridian's maintenance team has years of labeled examples pairing sensor readings with confirmed component failures, wants to group similar flight routes by passenger behavior without any predefined categories, and wants a gate-scheduling agent that learns by trial and adjustment through repeated simulated scheduling rounds. Which learning types correctly match these three needs, in order?
- Meridian's data-science team wants to combine sensor readings, weather data, maintenance-log text, crew-assignment history, and dozens of other variables into one predictive-maintenance model, but the model's performance gets worse, not better, as more variables are added. What CPMAI concept explains this, and what is the general remedy category?
- Meridian's data scientists are deciding which of 200 raw variables to feed into the predictive-maintenance model, including engineering new variables (such as 'average vibration change over the last 10 flight cycles') from raw sensor logs, and dropping variables that add no predictive value. Which CPMAI concept covers this activity as a whole?
- Meridian's generative-AI knowledge assistant needs to process thousands of pages of maintenance manuals and dispatch procedures so a language model can search and reason over them. A vendor proposes 'tokenization and vectorization' as the first processing step. What should the PM understand this step accomplishes?
- Meridian's PM is told the crew-scheduling model 'generalizes well' during testing but the maintenance director asks what that phrase actually means in plain terms for the project. Which explanation is correct?
- Meridian's data-science lead says, 'once we have a working algorithm, feature set, and enough labeled data, the project is basically done — deployment is just a formality.' As the AI project manager applying CPMAI's ML-fundamentals view, how should this claim be evaluated?
CPMAI Methodology · 82 questions
- Meridian Regional Airlines' predictive-maintenance pilot is underperforming, and investigation finds the sensor logs feeding it have gaps and inconsistent labeling across aircraft types. Before evaluating a different algorithm, what should the project team address first?
- A traditional-software project manager new to Meridian's customer-service virtual-assistant initiative insists on committing to a fixed go-live date and an exact accuracy target before any data has been reviewed. What key methodological difference is this PM overlooking?
- Meridian's finance team asks for a business case before approving budget for the loyalty-program fraud-detection initiative. The vendor's proposal promises "99% fraud detection" with no baseline comparison. What should the PM require before accepting this as a valid ROI justification?
- Meridian's cargo-and-passenger demand-forecasting team has eighteen months of historical booking data overall, but three regional routes launched only four months ago. What data-related risk should the PM flag before committing to a firm delivery date for route-level forecasts?
- A computer-vision ramp-safety proof of concept performs well on a curated set of 200 sample images. The vendor recommends moving straight to airport-wide production deployment. What pitfall should the PM recognize here?
- Meridian's gate-scheduling optimization model performs well in offline testing, but once live, the operations team reports it doesn't adequately account for last-minute weather diversions. What does this scenario illustrate?
- After the customer-service chatbot launches, a Meridian stakeholder asks when the project will be "finished" so the team can be reassigned elsewhere. Consistent with CPMAI's view of AI project lifecycles, how should the PM respond?
- A software vendor pitches Meridian's leadership on an "AI platform that eliminates the need for data scientists entirely" for the maintenance-manual knowledge assistant. What should the PM caution leadership about?
- Meridian's VP of Operations tells the board the crew-scheduling AI initiative will "eliminate all scheduling conflicts within 90 days," even though the PM knows Business Understanding hasn't even started. What is the PM's most important action?
- Meridian is choosing a delivery approach for the baggage computer-vision project, where model accuracy and labeled-data needs will only become clear once initial testing against real footage begins. Which approach best fits this uncertainty?
- Meridian's PMO wants to reuse its standard enterprise waterfall software-delivery template, unmodified, for the loyalty-program fraud-detection project. What is the key adaptation the PM needs to make for this data-centric AI project?
- A new PM at Meridian asks why the team should adopt CPMAI specifically, rather than continuing to run AI initiatives under the airline's existing generic enterprise project framework. What is the best answer?
- Meridian's dispatch-documentation generative-AI assistant has just completed its Business Understanding phase, with a scoped question, defined metrics, and a Go decision. Which phase comes next in the CPMAI methodology?
- Meridian's ops team tells the PM they want to "use AI for maintenance." What should the PM do first, during Business Understanding, to turn this into an actionable AI-specific business question?
- A Meridian dashboard shows raw sensor readings being turned into flagged fault patterns, then into a recommended maintenance action, and finally into a documented rationale a technician can trust and act on. Which CPMAI concept describes this progression?
- Meridian has five candidate AI initiatives but a limited Year One budget. Which scoping approach best reflects CPMAI Business Understanding practice?
- Meridian's team built a small offline demo of the fraud-detection model against a static sample dataset purely to test technical feasibility, with no plan yet to touch production data or live systems. What is this best classified as?
- Comparing two candidate Meridian projects — a chatbot that can reuse an existing pretrained language model, versus a from-scratch predictive-maintenance model requiring months of sensor-data labeling — what should the PM expect regarding time-to-ROI?
- During Business Understanding for the maintenance-manual knowledge assistant, the PM wants to speed up drafting candidate use-case questions and synthesizing stakeholder-interview notes. How can generative AI reasonably accelerate this phase of the project?
- Meridian's rebooking assistant needs to both interpret a passenger's free-text request (a cognitive task) and look up that passenger's itinerary in a legacy reservations database (a deterministic lookup). Why does separating these two components matter during Business Understanding?
- Meridian wants to auto-populate a maintenance ticket with a fixed set of fields whenever a technician submits a standard form, where the field mapping is rule-based and never changes. Should this capability be built as an AI/ML solution?
- Meridian wants a system that flags unusual mileage-redemption behavior in the frequent-flyer program — activity that deviates from a given member's normal pattern. Which of the seven AI patterns best matches this business need?
- Meridian is standing up the predictive-maintenance project team. Beyond a data scientist and an ML engineer, who is essential to include from Business Understanding onward to keep the project grounded in operational reality?
- A Meridian business analyst with strong spreadsheet and dashboarding skills wants to help with the demand-forecasting project but has no formal machine-learning training. How should the PM appropriately involve them?
- Before development begins on the fraud-detection model, the PM asks the business sponsor to define what "good enough" performance looks like. Why is establishing acceptable performance metrics during Business Understanding critical?
- At the end of Business Understanding for the crew-scheduling project, the team has a scoped question, defined metrics, and identified data sources — but has not yet confirmed the pilots' or mechanics' union will allow use of scheduling data for model training. What should the Go/No-Go assessment conclude?
- Meridian's finance team approved the ramp-safety computer-vision initiative after Business Understanding produced a scoped pilot targeting one terminal, a clearly defined false-positive tolerance, and confirmed camera-feed data availability. What does this scenario illustrate?
- Having received Go approval, the ramp-safety team moves into Data Understanding. What AI-specific question should they now prioritize that a traditional software project would not typically need to ask?
- Meridian Regional Airlines' predictive-maintenance team has moved from the Business Understanding phase into Data Understanding. The project lead reminds the team that this phase is not "just Business Understanding again with more data." What is the AI-specific focus that distinguishes the Data Understanding phase in CPMAI?
- Meridian's predictive-maintenance team has three candidate data sources for forecasting component wear on its regional jet fleet: five years of structured sensor telemetry with consistent timestamps, two years of maintenance technician notes in free-text form, and a vendor's generic industry failure-rate table that isn't specific to Meridian's aircraft. During Data Understanding, how should the team evaluate which datasets are "appropriate" for this ML initiative?
- Meridian's crew-scheduling optimization project has access to eighteen months of pilot and cabin-crew assignment history. The data science lead flags that this window barely covers one full seasonal cycle and includes no major weather-disruption events comparable to what the model must eventually handle. From a training-data-requirements standpoint, what should the project manager conclude?
- Meridian's loyalty-program fraud team has a labeled dataset of past mileage-redemption cases marked "fraudulent" or "legitimate" by fraud analysts. Before the model can be trusted, the project manager wants to validate this "ground truth." Which activity most directly validates ground-truth quality for this dataset?
- Meridian wants a demand-forecasting model for a brand-new route it just launched, but only has six weeks of bookings data — far too little history for a reliable seasonal forecast. The route is a strategic priority and leadership wants a working model soon. What is the most appropriate way to optimize this AI project given limited data availability?
- While assessing data for the ramp computer-vision safety project, Meridian's team discovers that the business question as originally scoped — "detect all ramp safety violations" — is far broader than any available camera footage can support; usable labeled footage only covers baggage-handling zones, not fueling or pushback areas. What should the project team do?
- Before Meridian's knowledge-assistant project (a generative-AI tool for maintenance manuals) moves from Data Understanding into Data Preparation, the project manager runs a phase-exit check confirming the manual corpus is complete, correctly scoped to the aircraft types in the fleet, and sufficiently current. What is this checkpoint an example of?
- Entering Data Preparation for the ramp computer-vision safety project, Meridian's team must formulate data-preparation requirements before any labeling work begins. Which of these belongs in that requirements-formulation step?
- Meridian's knowledge-assistant project has thousands of unstructured PDF maintenance manuals and dispatch procedures that need to be chunked, cleaned of scanning artifacts, and organized before the generative-AI system can be built on top of them. How can generative AI itself be applied to streamline this Data Preparation work?
- Meridian's ramp-safety computer-vision project needs thousands of video frames labeled for foreign-object-debris (FOD) detection, a highly specialized visual task requiring familiarity with ramp operations. A generic crowdsourced labeling vendor is cheap and fast but has no aviation-ramp experience. What should determine Meridian's data-labeling approach?
- Meridian's customer-service virtual-assistant team is preparing historical chat transcripts to train the rebooking-intent model. The raw transcripts include duplicate sessions from page-refresh bugs, agent internal shorthand notes accidentally logged, and inconsistent capitalization of city and airport names. What phase of work does resolving these issues represent?
- Meridian's cargo-demand-forecasting team has finished cleansing, labeling, and structuring its historical booking and cargo-volume data. Before committing engineering time to Model Development, the project manager wants to run the CPMAI Phase III Go/No-Go assessment. What should this checkpoint primarily evaluate?
- In a Phase III retrospective, Meridian's loyalty-fraud team describes how they spent three weeks reconciling mileage-redemption records across two legacy account systems, resolving conflicting customer-ID formats, and building a single unified feature table before any model training began. Which real-world CPMAI Phase III pattern does this best illustrate?
- While cleansing sensor data for the predictive-maintenance model, Meridian's data engineers find that roughly 8% of readings are missing due to intermittent sensor connectivity, and the failure events in the historical log are rare compared to normal-operation readings. Which pairing correctly matches each problem to the right data-preparation response?
- Meridian's project manager is choosing between an in-house team of ground-operations staff to label ramp-safety video (slower, higher domain accuracy, higher internal labor cost) and an external specialty vendor with aviation-safety labeling experience (faster turnaround, contractual quality guarantees, per-label fee). What is the correct project-management framing for this decision?
- Entering Model Development for the predictive-maintenance initiative, Meridian's project manager works with the data science lead to define the model-development requirements before any modeling begins. Which of the following belongs in that requirements definition?
- Meridian's data science team is under time pressure to deliver a first working version of the crew-scheduling optimization model. Rather than building and training a scheduling algorithm entirely from scratch, they start from an established open-source optimization library and adapt it to Meridian's duty-time rules. From a CPMAI Model Development standpoint, what does this represent?
- For Meridian's customer-service virtual assistant, the data science lead recommends starting from a pretrained large language model and fine-tuning it on Meridian's rebooking, baggage, and loyalty-program transcripts, rather than training a language model from scratch. As project manager, what is the strongest justification for approving this approach?
- Meridian's predictive-maintenance model has finished an initial training run. Before the team commits to a full Model Evaluation cycle, the project manager runs the CPMAI Phase IV Go/No-Go assessment. What should this checkpoint focus on?
- In a Phase IV retrospective on the cargo-demand-forecasting model, Meridian's data science team reports that their first three modeling approaches all technically "worked" in that they trained successfully, but each embedded an assumption (fixed pricing, no seasonal shocks, only one aircraft type) that made it unsuitable for Meridian's real cargo operation. What lesson does this real-world Phase IV example best illustrate?
- Meridian's predictive-maintenance model must ultimately be signed off by FAA-regulated maintenance technicians who are required to document their reasoning before grounding an aircraft component. A highly accurate but "black box" model is available, alongside a slightly less accurate model whose flagged predictions come with a clear explanation of which sensor readings drove the alert. Which factor should most heavily influence the model-development requirement here?
- Meridian's engineering team is deciding how to build the crew-scheduling optimization engine: license a mature commercial airline-scheduling optimization platform that already encodes common duty-time constraint patterns, or build a fully custom optimization model in-house from first principles. From a Model Development acceleration standpoint, what should most heavily inform this build-versus-buy decision?
- A new project coordinator at Meridian is confused about the difference between Data Understanding and Data Preparation, since both phases "deal with data." Which distinction correctly separates the two CPMAI phases?
- During labeling of ramp-safety footage, two ground-crew reviewers frequently disagree on whether a given object in frame counts as "foreign object debris" versus ordinary equipment left in a designated staging area. The project manager notices this disagreement is happening on a large share of ambiguous frames. What is the appropriate response?
- While cleansing and structuring the gate-scheduling dataset during Data Preparation, Meridian's engineers realize that a full third of historical gate-assignment records are missing the connecting-flight linkage field entirely — a field the original Data Understanding review had assumed was consistently populated. What should the project team do?
- For the loyalty-fraud model, Meridian has very few confirmed fraud cases compared to legitimate redemptions. A team member proposes using a generative-AI model to create additional synthetic fraud-pattern examples to help balance the training set, separate from the earlier step of cleaning up existing transcript records. How should the project manager categorize this proposal?
- Meridian's ramp-safety Phase III Go/No-Go review turns up a split: the prepared labeled dataset is technically clean and well-structured, but the mechanics' union has not yet been consulted about how flagged safety events will be used, and that consultation is still weeks away. Should the project proceed into Model Development?
- During Model Development requirements for the loyalty-fraud initiative, the business goal is to flag suspicious redemption behavior without a predefined list of known fraud types, since new fraud patterns keep emerging that historical labels don't cover. Which modeling approach should the project manager expect the data science team to propose, and why?
- Meridian Regional Airlines' data science team reports that their predictive-maintenance model hits 96% accuracy on held-out sensor data. Before the AI project manager signs off on the CPMAI Model Evaluation phase, what should the evaluation plan check first?
- Meridian's project manager is reviewing the evaluation plan for the crew-and-gate scheduling optimization model before it leaves the CPMAI Model Evaluation phase. Which addition would make the plan most comprehensive?
- Meridian's loyalty-program fraud detection model flags a large batch of legitimate frequent-flyer redemptions as suspicious during evaluation, frustrating customer-service agents who must manually clear them. What does the CPMAI model iteration process call for at this point?
- Six months after deployment, Meridian's cargo-and-passenger demand forecasting model starts under-predicting bookings on several routes. Investigation shows fuel-price surcharges and a competitor's new routes have shifted booking patterns since the model was trained. What is this scenario an example of?
- Meridian's customer-service virtual assistant continues receiving the same mix of rebooking, baggage, and loyalty questions it was trained on, but its accuracy at correctly escalating complex cases to a human agent has quietly declined over a year. No input patterns have changed. What should the project manager suspect is happening?
- Meridian's computer-vision ramp-safety model correctly detects foreign-object debris 94% of the time in evaluation testing, exceeding the technical accuracy target, but it misses two rare but severe debris types that have historically caused ground-crew injuries. What should the CPMAI Phase V (Model Evaluation) Go/No-Go assessment conclude?
- Which of the following best illustrates Meridian correctly executing the real-world activities of CPMAI's Phase V, Model Evaluation?
- Meridian's loyalty-fraud model reaches 98% technical accuracy, yet finance reports that total fraud losses on the frequent-flyer program have not measurably decreased since deployment. What does CPMAI's approach to Model Evaluation say the project team should do?
- During Model Evaluation, Meridian's team discovers the generative-AI knowledge assistant for maintenance manuals gives confidently wrong answers because the underlying document set it was built from was missing entire chapters of the newest aircraft's maintenance manual. What is the correct CPMAI response?
- After early testing shows Meridian's customer-service virtual assistant performing well on simple rebooking questions, an executive tells the board the chatbot will 'handle nearly all customer inquiries without human involvement.' The evaluation team's own criteria show it reliably handles only about 60% of real inquiry volume. What should the AI project manager do?
- Meridian's PM is preparing two separate Go/No-Go checkpoints for the demand-forecasting initiative: one at the end of Model Development, one at the end of Model Evaluation. What is the key difference in what each checkpoint should confirm?
- Meridian's team is building the generative-AI knowledge assistant for maintenance manuals. Rather than training a large language model from scratch, what approach does CPMAI's Model Development guidance point to as the practical accelerator?
- At the end of Model Development, Meridian's ramp-safety computer-vision model has been built and runs successfully on sample footage, but has not yet been tested against the formal evaluation criteria set earlier in the project. What is the appropriate Phase IV Go/No-Go outcome?
- Which activity best represents Meridian correctly executing CPMAI Phase IV, Model Development, rather than an adjacent phase?
- Meridian's predictive-maintenance model has just cleared its Phase V (Model Evaluation) Go/No-Go with a 'Go.' Correctly following the CPMAI methodology's six-phase order, what comes next?
- A newly hired project coordinator at Meridian assumes that once the demand-forecasting model reaches Operationalization, the CPMAI project is finished and the team can be fully reassigned. How does CPMAI's view of the AI project lifecycle correct this assumption?
- After the FAA updates duty-time regulations, Meridian's crew-and-gate scheduling optimization model — unchanged since deployment — begins producing schedules that violate the new rules, even though it never violated the old ones. What CPMAI concept does this best illustrate?
- Meridian's fraud analysts report that mileage-redemption fraud volume and type look statistically similar to a year ago, yet the unchanged fraud-detection model is now missing schemes it used to catch reliably, because fraudsters have subtly changed their tactics to exploit the model's specific blind spots. What is this an example of?
- Meridian's data science team wants to evaluate the ramp-safety computer-vision model using only offline metrics computed on a held-out image test set. What gap should the AI project manager point out in this evaluation plan?
- During Business Understanding, Meridian estimated a 9-month time-to-ROI for the crew-scheduling optimization initiative. Model Evaluation results now show the model performs well but only on Meridian's two largest hubs, not the smaller stations originally in scope. What should the project manager do with the original ROI estimate?
- A Meridian stakeholder asks the project manager to list the CPMAI methodology's six phases in the correct order. Which sequence is correct?
- A Meridian program office trained in traditional waterfall delivery wants to treat the Model Evaluation Go/No-Go as a single final sign-off, after which the project scope is permanently frozen and no further changes are allowed. Why does this clash with CPMAI methodology?
- During the Model Evaluation Go/No-Go meeting for the demand-forecasting model, data scientists present precision, recall, and mean absolute percentage error figures to a room of route-planning and finance executives who are unfamiliar with those terms. What is the AI project manager's most valuable role in this meeting?
- Meridian's evaluation team struggles to find enough real historical examples of rare, severe ramp-safety incidents to fully test the computer-vision model against. How might generative AI be used, consistent with CPMAI's view of generative AI as a project accelerator, to strengthen the Model Evaluation phase here?
- Meridian's crew-scheduling model fails its Phase V Go/No-Go because it underperforms specifically on holiday-period schedules, while performing well the rest of the year. What is the most appropriate CPMAI iteration response?
- During Business Understanding, Meridian's team established that the customer-service virtual assistant should be judged on successful resolution rate and appropriate escalation rate. At the Model Evaluation Go/No-Go, the data science team instead presents only model confidence scores and response latency. What governance gap should the project manager raise?
Machine Learning · 26 questions
- Meridian Regional Airlines' data science team is building a model that reads sensor readings from an aircraft component and outputs one of two labels: "likely to fail within 30 days" or "not likely to fail." The training data includes past component records already tagged with the outcome that actually occurred. Which type of algorithm is Meridian's PM overseeing here?
- Meridian's loyalty-fraud team combines the votes of several different models — a decision tree, a logistic model, and a gradient-boosted model — into a single fraud/not-fraud verdict for each mileage-redemption transaction, rather than relying on any one model alone. What technique is the team using, and why would a PM want it here?
- Meridian's network-planning team wants to group its ~120 routes into segments with similar demand-seasonality patterns, but they have no pre-existing labels defining which routes "should" belong together. Which type of algorithm fits this need?
- Meridian's crew-and-gate scheduling team is prototyping a system that repeatedly tries different gate assignments in a simulated version of a busy hub, receiving a numeric reward each time based on minimized taxi time, fewer connection misses, and no FAA duty-time violations, and gradually improving its assignment strategy over many simulated days. What kind of system is this, and what are its two core components a PM should be able to name?
- During a review of the gate-scheduling reinforcement learning project, a Meridian ops director asks why the data science team keeps letting the system try unusual, seemingly worse gate assignments instead of always picking the best-known assignment so far. What concept should the PM use to explain this design choice?
- Meridian's computer-vision team proposes a baggage-mishandling detector that achieves 97% overall accuracy in testing. The PM reviewing the results notices the model still misses a meaningful share of true mishandling events (false negatives), even though its accuracy score looks strong. What should the PM do?
- Meridian's virtual assistant needs to route each incoming customer message into one of four known intents — rebooking, baggage status, loyalty questions, or check-in — using thousands of past messages that support agents have already tagged with the correct intent. Which approach best fits, and why is it not clustering?
- At a Go/No-Go review for the ramp-safety camera system, a Meridian union representative asks the PM to explain, in plain terms, how the neural network decides a ground crew member has entered an unsafe zone. Which explanation correctly describes the basic structure the PM should reference?
- At a Go/No-Go review for the ramp-safety camera system, a Meridian union representative asks the PM to explain, in plain terms, how the neural network decides a ground crew member has entered an unsafe zone. Which explanation correctly describes the basic structure the PM should reference?
- Meridian's data science team recommends a neural network over a hand-written rules engine (e.g., "flag if pixel-brightness delta exceeds X near a marked zone") for the ramp-safety camera system. What is the core justification a PM should expect to hear?
- Meridian's PM notices the same general neural-network approach is being proposed both for cargo-demand forecasting (a numeric prediction) and for ramp-safety violation detection (a category decision). Is this a red flag, and why or why not?
- Meridian's dispatch team is asking the data science team to move beyond a shallow one-hidden-layer network to a deep architecture with many hidden layers for interpreting complex, multi-sensor engine-vibration patterns tied to developing faults. What is the PM-relevant justification for adding depth?
- Meridian's engineering team is comparing two deep learning approaches for two different projects: recognizing foreign-object debris in ramp camera footage, and modeling the trend in a time-series of turbine vibration readings over the past 90 days. Which pairing correctly matches architecture family to problem for a PM sizing the two efforts?
- Meridian has a small predictive-maintenance dataset — only 400 labeled failure events collected over several years for one aircraft type. The data science team proposes a very deep, high-capacity neural network. What should the PM raise as a concern?
- Meridian wants a system that maintenance and dispatch staff can ask natural-language questions of, drawing on the airline's maintenance manuals and dispatch procedures, and receive a synthesized, readable answer rather than a list of document links. Which technology is an appropriate fit, and why?
- During testing, Meridian's generative-AI dispatch assistant confidently states a specific regulatory requirement that turns out not to exist in any actual FAA or Meridian document. What should the PM take away from this, and what should the project do about it?
- A union representative at Meridian, unfamiliar with AI, asks the PM to explain "in plain terms" how the maintenance-manual chatbot actually produces its answers. Which explanation is both accurate and appropriate for the PM to give?
- Meridian's maintenance-manual assistant gives inconsistent answers to nearly identical questions from different technicians, and the data science team has not yet retrained or fine-tuned anything. What is the appropriate first step the PM should direct the team to try?
- After exhausting prompt-engineering improvements, Meridian's data science team still finds the general-purpose base LLM struggles with the airline's specific maintenance terminology and internal procedure codes. What is the appropriate next step, and what should the PM understand about it?
- For the dispatch-procedure assistant, the PM directs the team to design it so it always drafts a suggested answer with source citations for a human dispatcher to review and approve before any procedure change is acted on, rather than letting the model act autonomously. What principle does this design reflect?
- Meridian's customer-service chatbot, built on a general-purpose LLM, is asked by a passenger whether their specific flight departing in two hours is currently on time. The chatbot has no live connection to Meridian's operational systems. What limitation does this expose, and what does it imply for the project?
- During the training phase of the cargo-demand forecasting model, the PM's data science lead asks whether the PM needs to personally review each hyperparameter setting before training runs. What is the PM's appropriate role during this phase?
- To speed up development of the predictive-maintenance neural network, Meridian's data science team proposes starting from a pre-trained model already trained on general sensor/vibration data, then adapting it to Meridian's specific aircraft components, rather than training an entirely new network from scratch. What is this technique, and why does a PM care?
- Meridian's predictive-maintenance team uses one cloud vendor's ML platform, while the cargo-demand-forecasting team independently adopted a different open-source ML toolkit, and the two teams' outputs and data formats are incompatible with each other. What does this situation illustrate, and what is the PM's responsibility?
- After the platform-fragmentation issue surfaces, Meridian's PM proposes a shared feature store and a common data/model versioning environment so the predictive-maintenance and demand-forecasting teams stop working from disconnected spreadsheets and ad hoc scripts. What is this PM action an example of?
- Meridian is evaluating two cloud ML platform vendors for the ramp-safety computer-vision system, which needs GPU-accelerated real-time inference and the option to run inference at the edge (on-site, near the cameras) due to connectivity constraints on the ramp. One vendor is cheaper but offers only cloud-hosted batch inference with no edge deployment option; the other costs more but supports real-time GPU inference and edge deployment. What should guide the PM's platform decision?
Data for AI · 26 questions
- Meridian Regional Airlines' predictive maintenance lead tells the steering committee that the AI initiative's success depends far more on the sensor and maintenance-log data than on which vendor's algorithm the team ultimately picks. Which CPMAI principle does this statement reflect?
- Meridian's cargo-and-demand forecasting initiative ingests booking transactions, real-time weather feeds, and continuous engine-sensor telemetry from the in-service fleet, all arriving at different speeds and in different formats. Which Big Data characteristic is the PM primarily managing when comparing the batch-loaded booking data against the continuously streaming sensor telemetry?
- Meridian's computer-vision ramp-safety initiative has years of raw security-camera footage of the tarmac but no structured records of past foreign-object-debris incidents. Before this footage can train a detection model, what must the project team do to extract usable value from it?
- Two years into building a central repository for all of Meridian's flight, maintenance, and passenger data, the team discovers nobody can find or trust anything in it because data was ingested with no cataloging, quality checks, or access controls. What lesson from prior Big Data implementations does this illustrate?
- Meridian's analytics team currently produces dashboards that report last quarter's on-time performance (descriptive analytics). Leadership now wants to know which upcoming routes are likely to see demand spikes. Which shift in the role of data science does this represent?
- Meridian's engineering team needs to detect an abnormal vibration signature from an aircraft engine within seconds so ground crew can be alerted before the next flight, rather than discovering the anomaly in a report the next morning. Which Big Data approach should the PM push the team toward to enhance this AI capability?
- Meridian's predictive maintenance model needs both the structured numeric telemetry from engine sensors and the semi-structured free-text notes mechanics type into work orders describing what they observed. What does this combination best illustrate about data for AI at Meridian?
- Meridian's loyalty-program data — from initial mileage-transaction capture through model training, production scoring, and eventual archival once accounts close — needs a defined path so nothing is ingested, used, or retired inconsistently. What should the PM ensure is defined to manage this end to end?
- Meridian is standing up its computer-vision ramp-safety program and multiple teams — IT, safety, and ground operations — all touch the camera-footage data differently. Who should the PM ensure is explicitly assigned to own data quality, access rules, and definitions for this dataset?
- Before Meridian's crew-scheduling optimization project begins pulling data from HR, payroll, and FAA duty-time systems, the PM wants retention rules, quality standards, and access permissions agreed upon in writing rather than decided informally as issues arise. What should the PM establish?
- An auditor asks Meridian's AI team to show exactly how a raw sensor reading became a specific input feature used by the predictive-maintenance model, including every transformation it passed through. Which capability must the pipeline have provided for the team to answer this?
- A passenger's name and loyalty number appear slightly differently across Meridian's booking system, the loyalty database, and the customer-service chatbot's records, causing the fraud-detection model to sometimes miss that all three refer to the same person. Which practice should the PM push the team to implement to fix this?
- Meridian's data governance lead publishes a policy stating loyalty-program data must be accurate, complete, and consistent across systems, but six months later nobody has measured whether the data actually meets that policy. What is missing from Meridian's governance approach?
- Meridian's predictive-maintenance program needs engine-sensor readings to keep arriving continuously from in-service aircraft rather than being manually collected in periodic batches uploaded by ground crew. What should the data engineering team design to support this need?
- Meridian's data engineers are building the ingestion-to-model pipeline for the customer-service chatbot and want it optimized specifically for AI workloads rather than reused unmodified from an existing generic reporting-ETL pipeline. What distinguishes an AI-optimized pipeline from a standard business-reporting ETL pipeline in this context?
- Meridian's IT architecture currently couples its data storage tightly to one legacy reporting application, making it hard to add the AI workloads the ramp-safety computer-vision system needs without disrupting existing reports. Which data engineering principle should the PM push the team to apply to the AI infrastructure?
- Meridian's predictive-maintenance model is trained monthly on a large historical archive of sensor readings and maintenance outcomes, but in production it must score a live stream of current sensor readings within seconds to flag an at-risk component. What should the data engineering team build to support both needs?
- Meridian plans to triple its fleet size and add sensors to new aircraft types over the next three years, and leadership wants the predictive-maintenance data architecture to keep working without a redesign each time the fleet grows. What should the PM ensure the data architecture is designed for?
- A new engineer joins Meridian's AI team and struggles to understand how data flows from the engine-sensor feed through several transformation steps into the predictive-maintenance model, because the only documentation lives in the original architect's memory. What should the team have created to prevent this?
- Meridian's fraud-detection pipeline has been running for a year, and the PM wants the pipeline itself, not just the model's output, to surface a warning when the incoming data's format, volume, or statistical properties suddenly change from what the pipeline expects. What is this pipeline-level capability primarily protecting against?
- Meridian's early predictive-maintenance model performs poorly, and investigation reveals the training data included maintenance logs with inconsistent units, missing timestamps, and mislabeled failure events. Which principle explains why the model's poor performance traces directly back to this training data?
- Meridian's engine-sensor data comes from three different sensor vendors, each recording temperature in a different unit and rounding differently, causing the same physical reading to appear as different values depending on the source. What should the data preparation team do first to improve data quality?
- Meridian's finance team already runs a standard monthly data-cleaning process for its budget reports — removing duplicates and fixing obvious typos. The AI team is told to reuse that same process unmodified for preparing labeled training data for the fraud-detection model. What AI-specific data preparation need does this reuse overlook?
- Meridian's demand-forecasting model consistently mispredicts cargo volume on routes affected by severe weather, and the team realizes the training data never included historical weather conditions alongside booking and cargo records. What data preparation technique addresses this gap?
- Meridian's computer-vision model for detecting mishandled baggage performs well in the training environment but struggles on camera footage taken at unusual angles, in poor lighting, or in light rain, conditions underrepresented in the original training set. Which technique should the data preparation team apply to make the model more robust to these conditions?
- Meridian's loyalty-program fraud-detection dataset contains 50,000 legitimate transactions but only 40 confirmed fraud cases, and the resulting model almost never flags anything as fraud because it learned that predicting 'legitimate' is right 99.9% of the time. What should the data preparation team do to address this?
Managing AI · 16 questions
- Meridian's data science team has finished building a predictive-maintenance model that forecasts engine bearing wear from sensor data, and wants to move straight to production. The program manager insists on a formal quality assurance step first. What should that QA step center on?
- Meridian's predictive-maintenance model scores 98% accuracy on the training data used to build it, but when engineers apply it to sensor readings from aircraft it has never analyzed, its false-alarm rate spikes and it misses real bearing failures. What is the most likely explanation, and what should the PM do?
- A simple rules-based scoring model Meridian built to flag suspicious frequent-flyer mileage redemptions performs poorly on both the training data and new transactions, catching only the most obvious fraud cases. What does this pattern indicate, and what is the appropriate response?
- Meridian's cargo and passenger demand-forecasting model achieves strong statistical accuracy (low mean absolute error) in testing, but the finance team reports that pricing and capacity decisions based on its forecasts have not improved route profitability. What should the PM conclude?
- Meridian's computer-vision system for detecting foreign-object debris on the ramp reports 96% precision and 91% recall in lab testing. Operations leadership instead wants to know whether the system is reducing ramp safety incidents. What does this situation illustrate about model evaluation?
- After three months in production, evaluation data shows Meridian's customer-service virtual assistant correctly resolves rebooking requests but frequently mishandles baggage-status questions, escalating them to human agents far more often than necessary. What is the appropriate CPMAI-aligned response?
- Before deploying the loyalty-program fraud-detection model to production, Meridian's PM wants a quality assurance checklist that goes beyond a single accuracy number. Which combination best reflects a thorough QA process?
- Meridian's data science team has a modest, limited dataset of past crew and gate scheduling decisions to train an optimization model. The PM asks how the team should validate the model given this data constraint. What is the appropriate validation approach?
- Meridian's generative-AI knowledge assistant for maintenance manuals has completed fine-tuning and passed evaluation. The PM now needs to plan its move into production use by dispatch staff. What does the transition from training to inference primarily involve?
- Meridian's predictive-maintenance model has cleared evaluation and is ready to move toward production. The PM is developing an operationalization strategy. Which set of concerns most appropriately belongs in that strategy?
- Meridian is deciding where to deploy its computer-vision ramp-safety monitoring system, which must process live camera feeds with minimal delay to flag foreign-object debris before ground crew are exposed to danger, and which touches operational data Meridian is reluctant to send off-site. What deployment approach best fits these requirements?
- Traffic to Meridian's customer-service virtual assistant spikes sharply during weather disruptions, when rebooking volume can jump tenfold within an hour, then falls back to normal levels once flights resume. What deployment approach best matches this workload pattern?
- Meridian's PM is evaluating cloud-based machine learning services to host the cargo and passenger demand-forecasting model. Several vendor offerings are available. What is the most appropriate basis for selecting among them?
- Meridian's loyalty-program fraud-detection model is now live in production, continuously scoring mileage-redemption transactions. What is the PM's responsibility regarding data lifecycle management once the model is deployed?
- Six months after deployment, Meridian's data science team wants to release an improved version of the predictive-maintenance model that uses newly available sensor types. What should the PM ensure is in place before this new version replaces the current production model?
- Meridian recently upgraded the cameras used by its ramp-safety computer-vision system to a higher-resolution model. Detection performance has since quietly declined, and dispatch is now questioning whether the alerts can still be trusted. What should the PM's response be, framed within Managing AI production practices?
Trustworthy AI · 18 questions
- Meridian Regional Airlines' loyalty-fraud model was trained to flag suspicious mileage-redemption patterns. In production, it disproportionately freezes the accounts of infrequent flyers who redeem miles to book trips for family members, while rarely flagging the redemption behavior of frequent business travelers. No malicious actor is involved and no data was mishandled. Which Trustworthy AI concern does this scenario primarily illustrate?
- Meridian Regional Airlines is planning a computer-vision system that watches ramp areas for foreign-object debris and safety-zone violations, but the same camera feeds also capture ground-crew members constantly throughout their shift. The project sponsor asks how to keep the initiative ethically sound rather than addressing it only after complaints arise. What should the AI project manager do?
- Mechanics at Meridian Regional Airlines are refusing to engage with a new predictive-maintenance rollout because they believe the AI system will autonomously decide which aircraft are airworthy and remove them from their sign-off authority entirely. The project manager knows the system only forecasts likely component wear and still requires a certified mechanic's inspection and sign-off before any maintenance action. What is the best response?
- During pilot testing, Meridian Regional Airlines' customer-service virtual assistant is found to routinely fail to escalate distressed passengers, such as those stranded overnight with young children, to a human agent, instead looping them through standard rebooking prompts. The project team confirms this as a genuine ethical gap before broader rollout. What is the most appropriate mitigation strategy?
- Two objections surface about Meridian Regional Airlines' ramp computer-vision system: one employee worries the cameras are building a permanent behavioral profile of individual ground-crew members that could be used in disciplinary action, and another worries the system is secretly conscious and 'watching' them. Which pairing correctly separates a real concern from a misconception?
- Meridian Regional Airlines' customer-service virtual assistant currently logs a passenger's full frequent-flyer number, home address, and stored payment method on every interaction, even for simple questions like checking gate information. The AI project manager is asked to reduce the assistant's privacy exposure. What foundational privacy principle should guide the fix?
- Meridian Regional Airlines' demand-forecasting model uses booking data from European codeshare partners, including passengers based in the European Union. A passenger emails asking Meridian to delete their personal data from the forecasting system. What GDPR-related obligation must the AI project account for?
- Meridian Regional Airlines wants to share its loyalty-program dataset with an outside analytics firm to benchmark redemption patterns against industry norms. The dataset as it stands includes passenger names, frequent-flyer numbers, and home addresses alongside the redemption behavior the analytics firm actually needs. What should the AI project manager require before the data leaves Meridian?
- Meridian Regional Airlines wants to publish aggregate maintenance-sensor findings in an industry research paper. A data engineer proposes simply replacing each aircraft's tail number with a random code, while keeping every other field, including exact maintenance dates and specific route pairs, unchanged. The AI project manager is concerned this is not sufficient anonymization. Why is that concern valid?
- Security researchers demonstrate that Meridian Regional Airlines' ramp computer-vision system can be tricked into ignoring a piece of foreign-object debris on the tarmac by placing a small printed pattern near it, which causes the model to consistently misclassify the debris as harmless surface texture. What category of AI security risk does this demonstrate?
- Meridian Regional Airlines' generative-AI knowledge assistant, used by dispatch staff to query maintenance manuals, has been in production for six months with no additional oversight since launch. The AI project manager wants to establish ongoing protection against emerging risks such as prompt-injection attempts hidden inside uploaded documents. What should be implemented?
- Meridian Regional Airlines' pilot union asks how the new crew-scheduling optimizer weighs seniority, duty-time limits, and disruption history when assembling a roster, since the union must sign off before the tool touches live schedules. The vendor's underlying optimization method is a proprietary constraint-solving algorithm the union does not need to understand. What is the appropriate transparency approach?
- A frequent flyer whose loyalty account was frozen by Meridian Regional Airlines' fraud-detection model formally disputes the decision and requests to know exactly why the account was flagged, on what date, and by which model version. The project team discovers this information was never systematically captured. What Trustworthy AI capability is missing?
- An FAA auditor reviewing Meridian Regional Airlines' predictive-maintenance program asks how the model determines which components are flagged for early inspection. The vendor supplying the model considers its exact feature-weighting scheme a trade secret and resists disclosing it in full. How should the AI project manager balance the auditor's request with the vendor's intellectual-property concerns?
- Meridian Regional Airlines' finance leadership must approve continued capital investment in the cargo-and-passenger demand-forecasting model. They are not data scientists and have asked the AI project manager to explain why the model raised its forecast for a particular route by 18%. What is the most appropriate way to communicate this?
- Meridian Regional Airlines completed a legal review of data-privacy obligations for its European codeshare routes when the demand-forecasting project launched two years ago. Since then, EU data-privacy rules affecting AI systems have continued to evolve. The AI project manager is deciding how to handle this going forward. What is the appropriate practice?
- During a major weather disruption, Meridian Regional Airlines' automated rebooking-prioritization tool consistently places connecting-itinerary passengers from a specific low-fare booking channel at the back of the rebooking queue, even when their original disruption occurred first. A review finds this booking channel is disproportionately used by passengers from a particular region. What regulatory concern should the AI project manager treat as a priority before continuing to rely on this tool?
- Meridian Regional Airlines' generative-AI knowledge assistant occasionally produces a confident, plausible-sounding but incorrect summary of a maintenance procedure when the source manual is ambiguous or recently revised. Dispatch staff have started treating the assistant's answers as equivalent to the official manual. What should the AI project manager do to address this?
These questions are original practice material and are NOT actual exam questions or brain-dump content. All vendor marks are trademarks of their respective owners. This site is not affiliated with, endorsed by, or sponsored by the exam vendor.