A municipal utility has a fixed year-end budget and must choose between two proposals: a data-quality remediation initiative that cleans up inconsistent meter-reading records across legacy systems, or a higher-visibility generative AI chatbot pilot for customer billing questions that would rely on those same records. Which proposal should the utility prioritize?
Select an answer to reveal the explanation.
Short Explanation
A flashy chatbot built on messy meter data is basically a nice paint job on a car with no engine: it'll look good in the demo and then confuse customers the moment the numbers don't add up. Fixing the underlying data quality first is the less exciting choice, but it's what actually makes the chatbot trustworthy later. Foundations before facades, every time budget forces a tradeoff.
Full Explanation
Prioritization under a fixed budget requires ranking dependencies, not just visibility. In this scenario, the chatbot pilot is explicitly built to answer billing questions using the same meter-reading records that remediation would clean up, which makes data quality a prerequisite rather than a parallel workstream. Funding the chatbot first risks deploying a customer-facing tool that surfaces the very inconsistencies it was meant to resolve, damaging trust in the broader AI program. Choosing the chatbot for visibility and momentum mistakes short-term optics for a sound investment sequence, since a poorly performing pilot can undercut political support faster than a quieter foundational project builds it. Splitting the budget evenly sounds fair but under-resources both efforts, likely delivering a chatbot that still trips on bad data and a remediation effort too thin to finish. The claim that prompt design can compensate for inconsistent source data misunderstands how generative AI tools work: better prompts shape how a model expresses an answer, not the underlying accuracy of the records it retrieves. A scope caveat: this logic applies specifically because the two proposals share a dependency; unrelated initiatives would be prioritized on separate criteria. An operational check: trace which fields the chatbot pilot would actually query and confirm how many of those fields the remediation project touches.