Gradient Labs transforms financial services customer support with Claude

Gradient Labs is transforming customer support for financial institutions by building an AI-powered platform that follows standard operating procedures with the precision and reliability required in highly regulated environments.

With Claude, Gradient Labs achieves:

• 80-90% resolution rates for customer support automation, with customer satisfaction outcomes that are as high as 98%
• The flexibility to automate almost 80% of tasks within customer support and operations functions
• Profitable outcome-based pricing that enables teams to automate complex resolutions
• An AI Agent platform safe and compliant to work in regulated banks

AI for the most demanding regulatory environments

Dimitri Masin, along with Neal Lathia and Danai Anthoniou, recognized an emerging pattern across customer support automation while at Monzo. Companies reported 40-50% automation rates, but were only automating simple queries that required minimal time anyway.

This insight revealed the real opportunity: automating complex, time-consuming queries that create meaningful operational impact. Rather than building a generic solution, Gradient Labs focused specifically on financial services, where their deep understanding of regulatory requirements, like detecting vulnerable customers and preventing unauthorized financial advice, allowed them to build appropriate guardrails from the start.

Selecting Claude for human-like responses and superior reasoning

After evaluating multiple AI models, Gradient Labs chose Claude to power the majority of its platform. "Claude's ability to generate more natural, human-like, and authentic responses compared to other models," said Masin, made it their primary choice for response formulation, intent classification, and reasoning over their knowledge graph.

Claude handles the most critical functions for Gradient Labs’ platform: understanding customer queries within context and executing standard operating procedures (SOPs) by determining appropriate next steps. "Claude's proficiency in understanding customer queries within the context of their knowledge graph to provide accurate responses and its effectiveness in executing SOPs by determining the next appropriate steps and utilizing relevant tools," said Masin.

Delivering results through intelligent automation

Gradient Labs leverages Claude across several key capabilities:

• Intent classification and reasoning: Claude analyzes customer queries and determines the appropriate response path using Gradient Labs' knowledge graph
• Standard operating procedure execution: Claude follows complex multi-step processes required in financial services, determining next steps and utilizing relevant tools
• Response generation: Claude creates natural, human-like responses that maintain the quality and tone required for regulated financial institutions
• Knowledge graph integration: Claude understands customer queries within the context of Gradient Labs' comprehensive knowledge base to provide accurate, contextual responses

Transforming customer support quality and efficiency

The impact on Gradient Labs' clients has been substantial. Rather than focusing solely on automation percentages, the company prioritizes resolution rates for end-to-end query handling without human intervention. "We focus on achieving 80-90% resolution rates, which leads to real operational savings," said Masin.

This approach delivers meaningful value—while competitors might automate simple queries with minimal impact, Gradient Labs' high resolution rates translate to substantial time savings for financial institutions. Their outcome-based pricing model, which includes tiered pricing for more complex resolutions, aligns their incentives with customers' goals of achieving higher automation rates.

For customers, the experience feels seamless and professional. The AI handles complex queries with the precision and compliance awareness that regulated institutions require, while maintaining the conversational quality that customers expect from premium financial services.

Building the future of intelligent customer support

Gradient Labs envisions AI transforming customer support by handling repetitive tasks while enabling human agents to focus on complex, relationship-building work. "AI handles the repetitive and often tedious aspects of customer support, freeing human agents to focus on more complex issues and providing exceptional, above-and-beyond service, similar to the early days of Monzo," said Masin.

Gradient Labs continues refining its platform to handle complex scenarios while maintaining the regulatory compliance and quality standards that financial services demand. With Claude as their core AI foundation, they can tackle advanced customer operations use cases, such as money laundering and fraud investigations, transforming how financial services deliver customer support.