By implementing Real-time Agent Assist and Auto QA, the health services provider has reduced onboarding times, improved coaching, and increased conversion rates by 4 percentage points.
Clinicians from Signify Health visit millions of Americans in their homes each year to provide health evaluations and preventive screenings. Behind the scenes, the company’s contact center handles a constant flow of calls from members, from scheduling appointments to updating insurance information.
As operations scaled, Signify Health needed a better way to onboard new hires and identify coaching opportunities, so agents can manage calls with more clarity and confidence. The company turned to Observe.AI for a solution, implementing Real-time Agent Assist to give agents live prompts and guidance during calls.
“Prior to Agent Assist, onboarding took three weeks. We’d train new hires, then drop them on the floor with no real assistance,” says Amanda Massaro, Senior Leader, Quality and Training, at Signify Health. That lack of real-time support could not only impact performance—it also risks agent confidence and retention, which can drive up training costs and turnover.
“Now we can guide and prompt them through tough calls. You can hear it in their voices that they feel more at ease.”
With Observe.AI’s Real-Time Agent Assist, agents get adaptive, step-by-step guidance through smart scripts and dynamic prompts tailored to their skill level, call context, and customer sentiment. AI-powered alerts flag issues in the moment—like negative sentiment or fast speech—helping agents course-correct instantly. This real-time support boosts confidence, consistency, and performance across every interaction.
That confidence boost has made a significant difference for Signify Health. “We’ve been able to take the opportunities we see in Real-time Agent Assist and apply them directly in our coaching sessions,” says Massaro.
By using Observe.AI to correlate prompt usage to business outcomes, Signify Health increased its speed to proficiency—the percentage of new hires reaching an acceptable level of performance—from 70% to 82% in one year.
Before implementing Agent Assist, less than 1% of new hires received formal coaching due to manual processes. With real-time smart scripts, the team can now provide targeted guidance at scale.
Real-time Agent Assist doesn’t just help newcomers learn and get up to speed faster. It also improves how all agents handle member objections.
“Before, agents would flip through pages of rebuttals, just script reading, panicking a little,” explains Massaro. “Now they get real-time prompts that suggest the best approach [to an objection]. They sound a lot more confident.”
That has led to a higher conversion rate—the percentage of calls that result in a scheduled in-home visit. “We previously had about a 30% conversion rate. We’ve seen that upwards of 34% just in the short time that we’ve been using Agent Assist.”
Likewise, by adopting Observe.AI’s Auto QA, the quality team has gone from manually reviewing around 70,000 calls a year to analyzing about 65 million. This change enables them to identify patterns and insights that would have otherwise gone unnoticed.
“The amount of trends and insights we’ve gotten has been really phenomenal,” says Massaro.
Implementing Observe.AI was technically seamless, according to Massaro. While newer agents embraced the platform quickly—having used it from day one—encouraging adoption among more tenured agents required a more tailored approach.
“New hires adopted it quickly because they had it from day one,” she says. “But for agents who’ve been around longer, we’re having to show them how it can fit into their workflow.”
Fortunately, Real-time Agent Assist provided invaluable support to agents at every stage of their journey. With personalized, dynamic guidance that adapts to agent skill level, call context, and queue type, it enhances—not disrupts—existing workflows. Teams also have the flexibility to customize prompts, scripts, and alerts, ensuring the tool feels like an extension of the agent’s strengths, rather than a replacement.
Still, many experienced agents have started to embrace it. “One of our power users said recently, ‘I don’t know how I did life before Agent Assist.’ That’s the kind of thing I love to hear.”
Signify Health is currently exploring VoiceAI to automate high-volume, low-effort interactions—such as scheduling appointments—with the goal of streamlining operations while ensuring a seamless, intelligent handoff to live agents when needed.
By offloading routine tasks to AI agents that speak, think, and act like your best performers, Signify’s team can focus more time and energy on complex, high-touch conversations that require empathy, nuance, and human insight.
“It’s not about reducing headcount,” says Massaro. “It’s about giving our agents the space to focus on the harder conversations.”
When these more complex calls are routed to live agents, Real-time Agent Assist kicks in—equipping them with rich conversation context, dynamic prompts, smart scripts, and AI-powered alerts. This ensures agents can confidently take over from where VoiceAI left off, maintaining continuity and delivering a consistent, high-quality customer experience.
And across it all, Auto QA closes the loop—analyzing 100% of conversations (AI and human-led alike) to provide teams with the insights they need to continuously improve performance, coach with precision, and scale what works best across the organization.
.jpg)
.jpg)
.png)
.png)
.png)
