G2A has launched an AI support agent called Dave for sellers on its digital marketplace, now available in 180 countries. In its first 63 days of full deployment, the system handled more than 14,400 seller support cases and cut the share of tickets needing human intervention by 27% to 30%. The move targets a recurring problem for customer support teams in gaming: sudden surges in enquiries tied to major video game releases.
Dave handles about 10,000 customer conversations a week worldwide, according to G2A. The agent connects directly to transaction records and seller policies, letting it check payment status, orders, refunds, vouchers, and digital game keys while handling a case. If the system cannot resolve an issue, it passes the case to the seller with a summary of the exchange so customers never have to repeat themselves. This kind of automated support is a growing focus in AI for Customer Support, helping teams manage unpredictable volumes without sacrificing service quality.
Trial results
G2A reported a routing accuracy rate of 93.8% during the initial deployment. The system directs cases either to the relevant third-party seller or to G2A support when the issue relates to the marketplace itself. The company also posted a customer satisfaction score of 4.16 out of 5 for the automated process, with 74% of users rating it as easy or very easy. Average response quality reached 91.2%, and negative in-chat feedback was just 0.81%.
These figures suggest G2A is measuring the tool on both efficiency and customer response, not just ticket volume. Dave differs from a standard chatbot because it relies on verified transaction data rather than generative text alone. That approach reduces the risk of incorrect answers and lets the system handle more complex support issues, including software licences and seller-issued refunds.
AI strategy
Paweł Wróbel, Chief Growth Officer at G2A, said the tool was built to respond to demand spikes that cannot be met by hiring staff at short notice.
"As G2A.COM continues to expand globally, ensuring that sellers can provide the highest possible standard of customer support remains a key priority. Activity across the gaming ecosystem, including major game launches, can trigger sudden spikes in user inquiries within hours, making it inefficient to scale support simply by hiring more agents. Dave was created to deliver immediate, high-quality, multilingual assistance at scale. If a case requires full manual intervention by the seller, the seller and their support team receive a comprehensive analysis and summary of the case so users never have to repeat themselves. Dave is also another step in building an AI-native organization and environment. 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 the platform scale globally while delivering faster and more trusted user experiences," said Wróbel.
Every response from Dave is automatically evaluated across 13 quality and safety dimensions. During the first phase, the system resolved about 54% of seller cases on its own, with the rest routed for manual handling. Human agents still handle a substantial share of cases, but G2A sees the most immediate gains in triage, verification, and resolution of straightforward issues at scale-especially when blockbuster game launches trigger a rush of customer contact.
Why this matters for customer support professionals
Dave's early results show how transactional data can ground an AI agent and keep error rates low, a practical model for support teams that face sudden workload spikes from product launches. The system's ability to verify payment status, check digital keys, and hand off only the unresolved cases with a summary reduces the need for customers to repeat themselves-a common friction point in automated support. For teams weighing AI tools, the key takeaway is that accuracy and clean handoffs matter more than conversation volume, and anchoring the agent in real transaction data rather than relying on generative text alone can build trust and reduce escalations.
Your membership also unlocks:
