The advancement of AI in customer service presents an opportunity for transformation, allowing users to expect timely assistance without the frustration of long wait times. Unfortunately, this vision often doesn't materialize. Customers seeking support frequently encounter automated systems that seem more focused on directing them away from human assistance than on genuinely resolving their issues. If a customer is directed to an unhelpful FAQ instead of receiving the help they need, this disengaging experience might still be considered a success by the service provider. Such scenarios can lead to increased customer churn and present risks, particularly in sensitive sectors like gambling, where financial distress or responsible gambling practices may be overlooked.
Since its inception as Conduet in 2019, Raphie, located in New Jersey, has prioritized authentic resolution and what it terms "accountable automation" in its AI platform tailored for iGaming. Drawing from over one million actual industry interactions, Raphie employs its technology to comprehend account context and determine the necessary actions for each customer inquiry.
"A key tenet of Raphie is to eliminate the experience where a customer feels like they are talking to a robotic system that lacks understanding of their questions," said Justin Heath, co-founder and CEO.
The effectiveness of Raphie's approach is evidenced by a US sportsbook client that achieved over 90% customer satisfaction among users assisted by its automated services. This success, combined with Raphie’s industry knowledge and a focus on helping operators cut costs without reducing service quality, paves the way for its ambitious entry into European markets.
Raphie’s expansion into Europe stems from the recognition that generic AI systems often lack the capability to handle the specific inquiries typically generated by iGaming establishments. Reports indicate that about 90% of inquiries in iGaming require account-specific data. Consequently, when a customer inquires about the status of a withdrawal, they might just receive a broad answer regarding general payment policies. Properly addressing such queries might necessitate identifying whether a fraud review is in progress, asking for further documentation, or providing a concise payment timeline.
This limitation extends to questions surrounding bet settlements, bonuses, and game rounds, where generic systems may recognize a general query but lack the detailed industry knowledge needed to explain, for instance, why a parlay bet didn’t win.
Raphie counters this issue using what it describes as a "context layer," which integrates with operators’ systems for player account management, wagering, and customer service. This integration enables the retrieval of relevant information, allowing the platform to tailor responses according to the operator’s procedures.
Heath noted that operators sometimes initially underestimate the need for a specialized platform because they already employ generic AI tools within existing help desks. However, there can be a significant disparity between the promised effectiveness of these generic systems and their actual performance.
"Operators will claim they are using platforms like Salesforce or Zendesk with AI tools onboard. They might expect a 70% or 80% automation rate, but typically discover it only delivers 5% or 10%, as that’s the volume manageable without integrating wagering technology," Heath explained.
Raphie presents an alternative that does not require an all-or-nothing transition from human agents to automation. Its system features three distinct modes that can function independently or be blended to suit operators' preferences.
In Full Automation, Raphie’s AI agent manages supported interactions from start to finish, ensuring smooth transitions to human agents when necessary. The Co-Pilot mode assists human agents by suggesting responses that can be used directly or modified. Additionally, Raphie offers access to Specialist Human Agents located in Jersey City and the Philippines for customer service, risk management, and payment support.
“We focus on identifying where customers lose track of the conversation, rather than simply viewing those interactions as resolved," said Heath.
A practical example of Raphie’s approach can be observed with BetSaracen, an Arkansas sportsbook that implemented Co-Pilot over a year ago and added Full Automation shortly after. Within the initial eight weeks of Full Automation, the platform achieved a 77% rate of autonomous resolutions and cut response times by over 55%, while maintaining customer satisfaction at above 90%.
Importantly, Raphie doesn't equate resolution merely with the absence of human involvement; it factors in customer satisfaction, whether an agent's intervention was necessary, and if a player followed up on the same issue within 24 hours. The company conducts extensive quality assurance to monitor and recognize abandoned conversations that did not reach a proper conclusion.
Heath highlighted the negative implications of abandoned conversations, stating, "If a customer finds themselves going in circles or feeling unsatisfied, that’s a significant failure. We actively review where customers disengage before their inquiries have been fully addressed."
As Raphie's co-founder and COO, Alex Jones, pointed out, the way some customers engage with the AI reflects the quality of their experience.
"Many customers interact with our AI agent without realizing they are not conversing with a human. We receive positive interactions, with users even wishing Raphie a good night, indicating a natural dialogue experience," Jones noted.
For operators looking to adopt automation, Raphie outlines a systematic four-stage approach. It begins with evaluating the operator’s inquiries and workflows, followed by a shadow mode for testing responses without sending them to customers. Automation is then introduced in stages based on categories, culminating in continuous optimization as products and player behaviors evolve.
“This gradual approach allows operators to not only increase the volume of inquiries managed by Raphie but also expand the range of actions it can entail for resolving them. For instance, if a bet-settlement inquiry indicates that while an event is concluded, the market remains unsettled, Raphie can alert the trading team while keeping the player updated. It can even notify game providers about issues or unlock an account based on the operator’s criteria," said Heath.
Despite the growing capabilities of automation, Raphie acknowledges that certain inquiries necessitate human involvement. Interactions that fall outside approved parameters or escalate sensitive issues trigger a handoff to human agents. Heath illustrated this with a scenario where a customer discusses a pending withdrawal and indicates urgency regarding their financial situation, stating, "That type of flag is escalated directly to the appropriate team."
The balance that Raphie seeks between enhanced automation and necessary human oversight frames its concept of accountable automation. The focus is not merely on maximizing automated conversations but ensuring that all decision-making remains measurable, auditable, and operates within an operator-defined scope.
As the technology continues to advance, Raphie anticipates a future where AI can handle a wider array of inquiries and take necessary operational actions to resolve them. This would allow human teams to focus on more complex cases while still ensuring oversight when issues of judgment, player protection, or regulatory compliance arise.
Having tested its model in North America, Raphie is now looking to expand into Europe. This transition won't simply replicate their approach from one region to another; Raphie refers to Europe as a regulatory mosaic where understanding of payments, bonuses, and KYC queries might apply, but local regulations must be adapted for each market. Player protection guidelines, responsible gambling triggers, disclosure requirements, and audit protocols can vary significantly between jurisdictions.
European operators already have advanced customer service structures, supported by skilled teams who understand the industry's complexities. Nevertheless, increasing gambling taxes and other pressures are compelling these operators to rethink their cost structures and reassess how they deploy human resources alongside technology to boost efficiency and improve customer experience. Raphie claims its solution can cut customer support expenses by more than 60%.
Heath noted, "There’s a stronger motivation in European markets to adopt AI rapidly due to heightened cost pressures. Interest is evident among operators of all sizes, from small companies to the largest players globally."
Through its expansion in Europe, Raphie aims to assist iGaming operators in established markets in resolving more customer issues efficiently while ensuring that human expertise is reserved for cases that truly require it. For customers, this should signify a shift toward more definitive solutions and fewer frustrating back-and-forth exchanges.
