Gradient Labs CEO and Co-Founder, Dimitri Masin, featured in Interview Series Discussion
Dimitri Masin founds Gradient Labs for AI-driven autonomous customer support in regulated industries
Dimitri Masin, formerly a senior leader at Monzo Bank, has launched Gradient Labs - an AI startup dedicated to creating autonomous customer support agents tailored for regulated sectors like finance. Masin, in his new role as CEO and co-founder, aims to blend high-performance AI with stringent regulatory compliance.
After witnessing a significant technological shift with the debut of GPT-4, Masin and his team recognized the potential to automate up to 80% of manual, repetitive tasks through AI in early 2023. This breakthrough inspired the formation of Gradient Labs.
Having experienced hypergrowth at Monzo, Masin has incorporated lessons from that journey into Gradient Labs. Key learnings include balancing autonomy with guidance, ensuring top talent receives corresponding compensation, avoiding the unnecessary innovation of organizational fundamentals, and focusing on quality over novelty.
To build an AI agent that excels in regulated environments, Gradient Labs prioritized quality over speed, taking 14 months to develop their agent before releasing it. Their approach included maintaining a high-quality benchmark, focusing on workflows beyond simple question answering, and limiting tool exposure while promoting extended chain-of-thought reasoning.
Gradient Labs' Otto agent goes beyond answering straightforward questions and handles complex tasks effectively. This is achieved through the use of SOPs (Standard Operating Procedures) and extensive chain-of-thought reasoning architecture.
Masin defines "superhuman quality" in customer support as delivering service exceeding human capabilities. This is measured primarily through customer satisfaction scores, with Gradient Labs consistently recording high CSAT scores.
While Gradient Labs has deliberately distanced itself from being tied to a specific language model provider, this choice has yielded benefits in terms of performance and reliability for clients. The company's architecture utilizes multiple models to adapt and evolve over time, which can help minimize costs and improve quality.
Beyond chatbots, Gradient Labs aims to automate back-office processes. Despite the technical and operational challenges involved in this kind of automation, the company believes success hinges on tackling both simpler and complex processes - the latter posing particular obstacles due to the need for effective knowledge transfer.
In blending fast efficiency with stringent regulatory compliance, Gradient Labs takes a deliberate, consistent approach that respects regulation while still ensuring swift customer responses. However, the company is realistic about the limitations of AI, with high-stakes decision-making remaining in the domain of humans.
Looking ahead, Masin sees AI reshaping the customer experience for banking and finance in various ways over the next few years. He envisions the emergence of true omni-channel interaction, adaptive UIs, better unit economics, exceptionally supportive customer service, and a transformation of customer support from an inconvenience to a valuable service.
(For more information on Gradient Labs and their endeavors, please visit their website.)
Dimitri Masin, in his new business venture with Gradient Labs, seeks to combine AI technology with stringent regulatory compliance, aiming to automate up to 80% of manual, repetitive tasks in the finance industry, particularly in back-office processes. The company's long-term goal is to reshape the customer experience in banking and finance by introducing true omni-channel interaction, adaptive UIs, better unit economics, exceptionally supportive customer service, and transforming customer support from an inconvenience to a valuable service.
