Square Yards uses Gemini Enterprise Agent Platform and Gemini APIs with other Google Cloud services to run SuperAgent Pro, a multilingual conversational voice engine. The public case reports 200,000 automated calls per day, sub-700ms conversational response latency, 12+ Indian languages and 10 billion Gemini AI tokens consumed per month.
基线:The case says manual dialing limited an individual employee to roughly 150-200 connected calls per day, but does not provide a controlled comparison for the automated system or define the workload denominator.量化结果:已登记
Supported Indian languages基线未公开 → 12 languages, lower bound来源原样报告 · current state as described on the public case pageOfficial Google Cloud customer case says the platform handles 12+ Indian languages; language-level quality, traffic distribution and evaluation method are not disclosed.
Automated conversational calls基线未公开 → 200000 calls per day, reported scale来源原样报告 · current state as described on the public case pageOfficial Google Cloud customer case reports 200,000 automated calls daily; call definition, successful connection rate, period and independent validation are not disclosed. This is a current scale measure, not a before/after improvement.
Gemini AI token consumption基线未公开 → 10 billion tokens per month来源原样报告 · current state as described on the public case pageOfficial Google Cloud customer case reports consumption of 10 billion Gemini AI tokens per month; token accounting scope, model mix, cost and period definition are not disclosed.
Conversational response latency基线未公开 → 700 milliseconds, upper bound来源原样报告 · current state as described on the public case pageOfficial Google Cloud customer case reports sub-700ms response latency; percentile, measurement harness, network conditions and workload mix are not disclosed.
ROI:公开投入、持续成本、收益或测量周期不足,未计算。
边界:Vendor-authored Google Cloud case; no Gemini model/version or API model ID is named, so this case remains organization-level. The reported scale and latency are not an independent productivity or ROI evaluation; call outcomes, conversion quality, human handoff rate, cost, measurement period and model-specific contribution are not disclosed.