G2A.COM launched 'Dave', a autonomous customer support agent that resolved 14,400 seller support tickets during a 63-day trial period. Built to handle surges in inquiries during AAA game releases, Dave now operates 24/7 across 180 countries and manages roughly 10,000 conversations per week.
How Dave works differently from chatbots
Unlike standard chatbots, Dave connects directly to a seller's transaction history, instantly verifying payment statuses and seller policies. This allows it to resolve cases involving vouchers, game keys, software licenses, and refunds without needing to generate synthetic answers. By relying on verified data instead of generative text, it avoids the pitfalls common in many AI for Customer Support tools, effectively eliminating AI hallucinations and errors.
Operational metrics from 63 days of full deployment
The trial produced clear operational results. Dave managed roughly 10,000 conversations weekly, streamlining 15,000 tickets and cutting the volume requiring human support by 27-30%. Key metrics included:
- Ticket reduction: 15,000 tickets streamlined, with a 27-30% drop in human-dependent cases.
- Routing accuracy: 93.8% of cases were instantly directed to the correct seller or G2A support.
- User acceptance: A 4.16 out of 5 satisfaction score, with 74% of users rating the process as easy or very easy.
- Quality control: Responses scored 91.2% on average across 13 quality and safety dimensions, and only 0.81% received negative in-chat feedback.
Embedding AI into the company's core
Paweล Wrรณbel, CGO at G2A.COM, said the tool goes beyond handling ticket spikes. "AI is no longer just a productivity tool or a competitive advantage. The real advantage comes from embedding AI into the core of how a company operates. Dave is a practical example of that approach, helping us and sellers using platform to scale globally while delivering faster and more trusted user experiences."
Why this matters for customer support professionals
For support leaders, Dave demonstrates that autonomous agents linked to real-time transaction data can absorb a significant share of routine requests immediately. That cuts overall ticket volume and lets human agents focus on cases requiring judgment. As AI Agents & Automation become more practical, teams can use similar systems to maintain service levels during demand spikes without expanding headcount. The agent's ability to provide multilingual, 24/7 support and send detailed summaries for escalated cases also reduces repetition and improves resolution time.
Your membership also unlocks:
