How Smarsh built an AI front door for regulated industries — and drove 59% self-service adoption

Key takeaways

• Smarsh successfully deployed "Archie," an AI-driven support agent, achieving a 59% self-service adoption rate among its customers.
• The solution, built on Salesforce's Agentforce 360 Platform, serves as an intelligent "front door" to centralize customer support and simplify navigation for regulated industries.
• Proactive data preparation over nearly five years was crucial for Smarsh to ensure clean, secure data, mitigating risks like AI hallucination and enabling rapid deployment.
• Smarsh anticipates significant improvements, including a 20% increase in self-service success, 25% faster issue resolution, and a 30% boost in service representative productivity.
• Addressing stringent regulatory compliance, data security, and Model Risk Management (MRM) requirements was central to Archie's development and adoption strategy.

What happened

Smarsh, a global provider of cloud-native, AI-driven solutions for capturing, archiving, and analyzing communications data in highly regulated sectors, embarked on an ambitious goal: to scale its workforce and boost productivity by 30% using AI. The company's customer service team identified a core challenge: customers faced a complex array of products, documentation, and compliance rules, making traditional support difficult.

To address this, Smarsh developed an intelligent, human-centric "front door" for customer support, named "Archie." This AI support agent was designed to simplify the customer journey by allowing users to describe their needs in plain language, bypassing complex navigation trees. Archie centralizes the support experience, directing customers to the right solutions efficiently.

Built on Salesforce’s Agentforce 360 Platform, Archie transitioned from concept to production, offering smart self-service and faster resolutions. While initial customer adoption required adjustments, Smarsh's strategic focus on personalization and improved change management ultimately led to a 59% self-service adoption rate.

Why it matters

Smarsh operates within highly sensitive, regulated industries, where data integrity, security, and compliance are non-negotiable. The successful implementation of Archie demonstrates how advanced AI technologies, specifically generative and agentic AI, can be effectively deployed in environments with stringent regulatory oversight, such as financial institutions.

This initiative is significant because it provides a blueprint for other companies navigating the complexities of AI adoption in regulated spaces. By prioritizing clean data, partnering with a trusted platform like Salesforce, and rigorously addressing infosec requirements and Model Risk Management (MRM) approvals, Smarsh has shown that AI can enhance customer service and operational efficiency without compromising regulatory commitments. The high self-service adoption rate underscores the potential for substantial improvements in customer experience and operational cost savings within these critical sectors.

Key details / context

Smarsh's journey began with a clear objective: leverage AI to improve workforce scale and productivity. Recognizing the "maze" customers faced, the company focused on creating a unified, intelligent support entry point. They chose Salesforce's Agentforce 360 Platform, integrating it with existing Salesforce products like Data 360, Agentforce Service, and Agentforce Sales, to ensure Archie could execute tasks across various systems automatically.

A critical factor in Archie's success was Smarsh's foresight in data preparation. Nearly half a decade ago, the company began meticulously rationalizing, annotating, and anonymizing its data, making it "clean and locked down" and ready for AI integration. This proactive approach significantly reduced the risk of AI "hallucination" and project stalls often seen in other initiatives.

For regulatory compliance, Smarsh went beyond standard measures. They merged their documentation team with the AI team, ensuring all material produced was verified and suitable for the Large Language Model (LLM). Discussions with Salesforce were crucial for obtaining MRM approval, satisfying banks and regulatory bodies about data protection and model transparency. Although initial rollout saw customer confusion regarding the new interface, a shift to emphasizing personalization and clear usage instructions boosted adoption to 59%.

What happens next

Smarsh expects the positive trend in customer self-service adoption to continue as Archie is progressively rolled out across its full suite of products. The ongoing partnership with Salesforce will facilitate the evolution of Archie and its underlying AI capabilities, adapting to new technological advancements as they emerge.

Internally, Smarsh will maintain its rigorous focus on data trust and security. The integrated documentation and AI teams will continue their tight collaboration, ensuring that all information fed into Archie remains pristine, verified, and compliant with regulatory requirements. This continuous vigilance underscores Smarsh's commitment to delivering reliable and secure AI-driven support in a highly regulated environment.

FAQ

Q: What is Smarsh's "AI front door"? A: Smarsh's "AI front door" is an intelligent, human-centric entry point for customer support, named "Archie." It allows customers to describe their needs in plain language, simplifying access to solutions.

Q: What platform did Smarsh use to build Archie? A: Smarsh built Archie using Salesforce's Agentforce 360 Platform, leveraging its shared context, controlled execution, and orchestration capabilities.

Q: How did Smarsh ensure AI compliance in regulated industries? A: Smarsh prepared clean, secure data over many years, integrated its documentation and AI teams, and worked with Salesforce to achieve Model Risk Management (MRM) approval, satisfying stringent data protection rules and regulatory bodies.