Uploaded February 2025 | Updated September 2026, 2 weeks ago
In this talk, AT&T showed how fine-tuning small language models (SLMs) has disrupted their call center operations, saving $2 million annually while outperforming expensive alternatives like OpenAI when balancing cost, latency, and accuracy.
This is the ultimate David vs. Goliath story in AI. By fine-tuning H2O.ai’s Danube, AT&T classified 80+ customer-agent interaction categories with precision, slashing costs by 85% and speeding up response times. It’s faster, cheaper, and a game-changer for efficiency.
In this talk, AT&T showed how fine-tuning small language models (SLMs) has disrupted their call center operations, saving $2 million annually while outperforming expensive alternatives like OpenAI when balancing cost, latency, and accuracy.
This is the ultimate David vs. Goliath story in AI. By fine-tuning H2O.ai’s Danube, AT&T classified 80+ customer-agent interaction categories with precision, slashing costs by 85% and speeding up response times. It’s faster, cheaper, and a game-changer for efficiency.
