When Manual Triage Becomes a Bottleneck
The client is a leading global financial technology enterprise offering products that simplify personal finance, accounting and tax preparation for individuals and businesses worldwide. It has long been a pioneer in financial management software across tax filing, accounting and business finance.
As the customer base expanded, service operations began to feel the strain. Thousands of tax and accounting queries arrived daily across email and phone. Each case had to be reviewed, interpreted and manually routed to the right tax analyst.
Over time, a familiar pattern emerged: manual triage slowed case assignment, SLA breaches increased, some queues were perpetually overloaded while others had spare capacity, analysts faced growing backlogs and fatigue and customers felt the impact in delayed responses and unresolved issues.
The client needed more than another point automation. It wanted a Generative AI (GenAI)–driven system that could intelligently automate case triage, route work based on skills and availability and improve customer responsiveness at scale.
From Manual Routing to Agentic Mesh
Persistent collaborated with UiPath to design an agentic mesh—a self-orchestrating network of AI agents that manage end-to-end case handling.
The design combined:
- Salesforce Agentforce agents to understand cases, map them to the right analyst skills and assign work accordingly
- UiPath Maestro agents to monitor queues, track load and rebalance cases as conditions change through the day
Instead of a single automation running in one platform, Persistent configured agents on Salesforce and UiPath to communicate via API calls and an Agent-to-Agent Protocol. Each agent could interpret the others’ inputs, outputs and task status. This made cross-platform hand-offs seamless and reduced the need for manual queue coordination.
In practice, agents on Salesforce and UiPath now operate as a unified system, not as isolated tools.
How a Single Email Finds the Right Expert
The value of the mesh is easiest to see in a simple, everyday interaction.
A customer sends an email saying they cannot enable the “Make Recurring” feature for payments.
- A UiPath agent reads the email, interprets it as a payment-related issue and classifies the case accordingly
- It notifies the Salesforce agent, which consults a live map of analyst profiles, skills and time zones to identify the best-fit analyst who is currently available
- As queues grow or shrink, the Rebalancing Agent dynamically redistributes cases across teams to avoid bottlenecks
From intake to resolution, the process is largely autonomous and aligned with SLAs by design. GenAI didn’t replace analysts; it made sure every case reached the right one, at the right time.
Numbers that Changed the Conversation
The agentic mesh delivered measurable gains across operations, productivity and experience.
Operationally:
- Approximately 50% reduction in SLA breaches, improving adherence to committed response times
- Backlog of delayed or unassigned cases reduced by nearly 30%, easing queue congestion and smoothing daily workflows
- Automated triage and assignment meant less time lost in hand-offs and more time spent resolving issues
On the productivity front:
- Analyst output improved by 10–15%, driven by more precise skill-based allocation, better use of time zones and continuous queue rebalancing
- Analysts focus on complex case resolution instead of manual sorting or status chasing
Service quality and workforce impact:
- Cases went to the most qualified analysts and human error in assignment dropped, improving first-time resolution and customer satisfaction
- Workforce managers were freed from constant queue firefighting and could focus on process improvement and experience design
- Fairer workload distribution across shifts and regions helped reduce fatigue and attrition risk
Template for Cross-Platform Service Intelligence
Strategically, this engagement showed that agents built on different ecosystems can operate as a single, coordinated layer of service intelligence. Salesforce Agentforce and UiPath Maestro are no longer isolated platforms; they form a unified fabric for routing, monitoring and balancing work.
This shift moves the organization from point automations to intelligent orchestration—where AI not only assists individual tasks but governs how work flows across the entire support landscape. The design has been validated as scalable and suitable for rollout across additional service lines and back-office operations.
The client chose Persistent for its cross-platform expertise (UiPath, Salesforce, Microsoft, Google Cloud), its early leadership in agentic deployments where multi-platform agents collaborate autonomously and its end-to-end engineering capabilities in automation, AI orchestration and enterprise integration.
Ready to move from manual triage to intelligent, cross-platform routing? Talk to Persistent.
