C_BAIPA sample questions & answers

You are the solution architect advising Nordhavn Freight, a Danish logistics group running ports and road haulage across the Nordics. Its integration estate runs on an on-premise middleware platform whose vendor support ends in 16 months; the vendor will extend support by six months at most, at a premium. The CIO has mandated SAP Integration Suite as the single target integration capability for all new and migrated flows.

An inventory lists 142 interfaces. The integration lead notes that many were built project by project: at least 35 move the same shipment-status data between the same systems through separate point-to-point flows, and a separate group of about 20 serve applications scheduled for decommissioning within the year. The team has four integration developers, and the CFO has capped the migration budget at a level the integration lead estimates covers about 90 interface rebuilds.

Three proposals are on the table. The integration lead favours a one-for-one rebuild of every flow so nothing is missed. The infrastructure manager suggests negotiating extended middleware support and moving flows only as projects touch them. An enterprise architect wants to apply the Integration Solution Advisory Methodology before committing to any migration plan.

Which recommendation best meets the support deadline, the budget cap, and the CIO's goal of a governed integration estate?

Seguros Caravela, a Brazilian property and casualty insurer, is piloting an AI agent built with Joule Studio to support its claims adjusters. The agent reads a claim, checks it against the policyholder's contract terms, and drafts a settlement recommendation that an adjuster reviews before anything is paid.

After three weeks the claims director reports mixed results. The drafts are fluent and correctly structured, and the agent retrieves claim headers and loss details from the claims system without errors. However, in 11 of 40 sampled cases the recommendation applied general market practice on deductibles and exclusions instead of the clauses in Seguros Caravela's own policy wordings, which differ by product line and issue year. Adjusters had to rewrite those drafts.

A configuration review confirms that the agent's instructions tell it to verify coverage against the applicable policy and that the claims-system connection is healthy. The agent's configured tools are a claims-system lookup and a recommendation-document template. Policy wordings are maintained by the underwriting department as versioned documents per product line and issue year, and underwriting issues revised wordings for each new issue year. The claims director wants the defect fixed before the pilot expands to two more regions, without exposing more personal data than the pilot already uses.

As the solution architect, which corrective action addresses the root cause of the defective recommendations?

Kestrel Ridge Water, a Canadian municipal water utility, is finalising the commercial plan for its SAP Business AI Platform programme. The programme has two parts. The first is a set of integration flows and extension applications that will run steadily from go-live, with volumes the utility can forecast from five years of meter and work-order history. The second is an AI agent pilot for field-crew scheduling whose usage could range from a dozen supervisors to all 600 field staff within the first year, depending on pilot results.

The account team has offered two commercial options that can be used separately or together. Under the fixed subscription, named services are bought at agreed quantities for the contract year; this offer allows no mid-year quantity changes, which are made only at renewal. Under the usage-based option, services draw down a pre-committed credit balance as they are consumed, and credits can be topped up at any point in the contract year at the committed rate. The quote prices the steady workloads lower under fixed subscription than under usage-based credits at forecast volumes. After the steady workloads, the approved envelope leaves room for pilot usage equivalent to about 150 users.

The CFO requires annual platform spend to stay inside that envelope. The CIO does not want commercial terms to hold the pilot back if adoption grows mid-year. The finance team prefers a single model to keep administration simple.

Which commercial recommendation best balances the CFO's spending envelope with the CIO's need to scale the agent pilot?

Aceros del Bajío, a Mexican structural-steel fabricator running SAP S/4HANA Cloud, receives roughly 400 incoming steel coils a week. Inspectors at its three plants will record coil test results on tablets during each shift. A fit-to-standard workshop confirmed that the standard inspection apps cannot capture the plant-specific test sequence, and the architecture board has approved a side-by-side extension on SAP Business AI Platform that consumes only released APIs, in line with the company's clean-core commitment.

Three requirements remain for the build plan. The inspection rule set is owned by the quality team, changes monthly, and must be updated by that team without waiting for a developer release. When a test fails, the ERP system must receive a quality notification and the affected batch must be blocked from production use. Results are posted throughout each shift, and coil stock is drawn for production continuously during every shift. Each plant has one quality manager, who works day shifts only. The ERP system exposes released APIs for creating quality notifications and for changing batch status.

The lead developer has drafted four three-part build plans and asks you, as the solution architect, to approve the one that meets every requirement.

Which build plan should you approve?

Rheinmark Pharma Logistik, a German pharmaceutical wholesaler, wants AI-supported demand sensing in place before the winter respiratory season, which starts in 12 weeks. The model needs SAP sales-order and inventory history from the ERP system plus two external feeds: regional pharmacy sell-out data from a market-data provider and public health surveillance indicators.

The Chief Data Officer has made SAP Business Data Cloud the company's data foundation and asked you to apply the Data and Analytics Advisory Methodology. SAP Business Data Cloud is already in place, with the ERP sales and inventory data available from an earlier finance project. The data governance lead proposes first cataloguing the sales and supply-chain data domains with complete lineage before any AI use case starts; the data team estimates 16 weeks. The head of supply chain counters that the season cannot wait, and the supply-chain data scientists offer to extract the ERP tables into a standalone notebook environment within two weeks.

None of the pilot data is patient-level; the sell-out feed is aggregated by region. Its licence restricts redistribution outside the company.

Which recommendation best meets the 12-week deadline while keeping the data architecture on its intended long-term path?

Straits Meridian Bank in Singapore has piloted an AI agent, built with Joule Studio and monitored through SAP AI Agent Hub, that reviews SME customers' cash-flow data and proposes changes to their overdraft limits. During the pilot, proposals reached relationship managers as emailed spreadsheets, and approved changes were re-keyed manually into the lending system, which accounted for most of the two-day turnaround. For production, the business sponsor wants the agent to apply limit changes directly, cutting turnaround from two days to minutes, and the agent's technical user has already been granted the write authorization to do so.

The model-risk team has two findings. First, the bank's AI policy, aligned with its regulator's expectations, requires a named human to remain accountable for any decision that changes a customer's credit exposure. Second, a sample review found the agent's proposals well reasoned in 94% of cases; in the rest it misread seasonal revenue dips as distress and proposed reductions.

The business sponsor argues that 94% accuracy plus full logging of every agent action is sufficient oversight. The CIO wants a production design the bank can extend to other credit products without re-architecting, and the head of SME banking wants turnaround to improve noticeably for customers.

Which production design should you recommend?