
September 2023 Releases
You've built and tested your AI Agent, and you’re ready to go live.
Now comes the next step in developing fully optimized AI Agents: understanding how it’s performing.
Unlike human agents, who might catch you up on their day’s performance in an informal check-in, AI Agents communicate performance through data alone.
That’s why having the right data, where you need it, and when you need it, is essential for monitoring AI Agents—and ultimately the difference between an AI Agent that meets targets and one that excels beyond them.
In this guide, we’ll help you monitor every element of your agent in Regal—from the predictability of its prompt logic, to the customer’s experience, to the KPIs it drives for your business.
We’ll touch on every layer of Regal’s reporting engine:
With these monitoring layers in place, you can deploy Regal AI agents with confidence and optimize them with ease.
When you're running AI agents that handle hundreds or thousands of conversations daily, you need immediate visibility into performance. Without it, you run the risk of the agent mishandling objections, providing inaccurate answers, and degrading the customer experience on a large volume of live calls.
Whether you've just deployed your agent for the first time or are tweaking a long-standing agent, Regal’s cross-platform live metrics (updated intraday, every 10 seconds) give you the visibility that you need to monitor conversation volume and quality in real time—to take immediate action to ensure new agents are performing to grade, and existing agents aren’t causing regression.
Whether you’ve just launched an AI agent or are fine-tuning performance mid-day, Regal Live provides you with an overview of all things live stats across your agents, campaigns, queues, and journeys.
This is your source of truth for understanding how things are going right now: You can catch regressions or unexpected behaviors before they impact performance at scale, make live decisions on where priority changes need to be made, and instantly know which agents can safely be handed more volume.
You can use the live dashboard to compare completed tasks and number of conversations across all active agents, and confirm that agents are actively dialing, handling conversations, and driving the intended outcomes (i.e., transferring qualified leads to licensed agents).
With Regal Live, you can answer critical questions like:
“Is my qualification AI Agent on track to reach 1,000 connected leads by end of day?”
“What percent of sales conversations are dropping off early, and why?”
“Is introducing AI Agents decreasing speed-to-dial and inbound call abandonment rate?”
The Agents page lets you dive deeper into agent-specific metrics to identify your top-performing AI agents, spot where underperforming agents need adjustments, and validate that your agents are driving helpful outcomes on a call-by-call level.
Some examples of what you can monitor on this page:
Fully understanding your AI agent's performance requires deep analysis, historical tracking, and the ability to tie metrics to your business outcomes.
Regal provides the comprehensive data and analytics needed to know whether your agents are helping you successfully meet and scale your KPIs.
With it, you can answer questions like:
For each AI agent you build, you'll automatically get access to a curated Reporting dashboard that tracks the metrics that precisely reflect how you measure and track towards key KPIs.
Regal’s implementation team works with you to tailor these dashboards to your specific use case and success metrics.
These dashboards allow you to track high-level KPIs like:
This Reporting does not just give you a static number for your top KPIs, it’s also designed to identify trends, while letting you drill-down to surface distinct takeaways quickly:
Regal's Reporting capabilities allow you to go beyond KPIs to uncover specific topical and performance patterns, as well as the conversational signals driving performance.
While standard performance metrics tell you what happened, metrics that are tailored to your agent’s specific use case, industry, and goals help point to why it happened—unlocking the ability to make surgical improvements.
In Regal, you can track custom, unique data points across calls to identify topical and sentiment related patterns, and surface the in-call data points that drive those patterns.
With Custom AI Analysis, you define the unique data points that get extracted from AI Agent conversations—specific objections, compliance language, product or competitor mentions. Regal automatically analyzes post-call transcripts using an LLM to capture this data.
This leaves you with both in-aggregate and call-level insights that are fully tailored to your use case, your language, and your unique call outcomes. So you know exactly what’s working, and where to prioritize scale and needed updates.
For example, with an inbound lead qualification agent, you may care not only whether a call was transferred, but also:
These Custom AI Analysis data points are accessible in Reporting, so you can monitor them in-aggregate, and identify patterns down to precise data points instantly.
Define your Custom AI Analysis variables to track patterns in conversation topics, identify common dropoff points in your flow, watch for concerns (AI Agent compliance adherence, contact reluctance to AI interaction, objections raised), and more.
The above monitoring layers are powerful—if you know what to look for. But what if you don’t? What if your AI Agent is failing for reasons you haven’t thought to track? What if critical patterns are buried under thousands of call transcripts?
That’s where Regal Improve comes in.
Regal Improve uses LLMs and unsupervised clustering to automatically extract the most relevant moments from calls and group them into topics and subtopics.
With no setup required, it identifies patterns across your AI agent conversations, helping you spot what's working and where gaps exist, and helping you prioritize and know where to drive improvement in your agent builds.
Regal Improve automatically categorizes conversations into topics like pricing objections, product questions, or technical issues, with further detail in sub-topic breakdowns.
For example, when monitoring competitor mentions, you could:
Within each topic and sub-topic, you can review associated key metrics (talk time, sentiment) to understand which topics are resulting in positive versus negative call outcomes.
With real-time data, historical reporting, and tailored insights, Regal provides comprehensive monitoring that turns every conversation into an opportunity for improvement.
This means that not only can you feel confident that you’re on top of every AI Agent deployment and tweak, but you’ll also have the right data to continuously optimize your agents.
Ready to build AI Agents that can reach their full potential? Get started with a Regal demo now.
Ready to see Regal in action?
Book a personalized demo.