Octave Therapy doubles inbound capacity with AI voice assistant

About Octave

Octave is a leading mental health provider serving thousands of clients through their network of professional therapists, operating across multiple systems and platforms to deliver accessible mental healthcare.
‍

Executive Summary

In 2024, Octave faced a critical challenge: potential clients interested in therapy services couldn't get immediate responses to their inquiries, especially outside business hours. This meant lost opportunities to support people at their moment of need.

After evaluating multiple solutions, Octave partnered with Hello Patient to deploy Mia, an AI voice assistant. Within two weeks, they had a functional prototype. Four months later, they've nearly doubled their capacity to answer inquiry calls and created an entirely new channel for client engagement through automated text conversations.

The Challenge

Raul Estrada, VP of Operations at Octave, knew his team had a problem. Clients seeking mental health support often reached out when they needed help most—evenings, weekends, and other times when staff weren't available.

"We were looking for ways to provide immediate responses and information for clients interested in our services, regardless of the time of day or how busy our team was," Raul explains. "Without a solution that could immediately answer any client inquiry, we were not able to support potential clients interested in therapy services with our providers."

The stakes were high. Every missed call represented someone seeking mental health support who might not call back.

Octave needed a solution that could:

• Respond instantly to client inquiries 24/7
• Maintain their high-quality standards for client experience
• Integrate with their existing multi-system infrastructure
• Deploy quickly without extensive development resources

Why Hello Patient

Raul's first encounter with Hello Patient came through a colleague who had worked with their CEO. But it was the product demo that sealed the deal.

"I visited their website where I was offered the option to test their product via a phone call," Raul recalls. "I was immediately impressed with the low latency and natural-sounding voice on the other side of the line."

After exploring alternative solutions that fell short on quality, Octave chose Hello Patient for three critical reasons:

1. Superior product quality
"Having a low latency NLP solution that could ingest our knowledge base and help answer basic questions was crucial for us while keeping our high-quality bar for client experience.”

‍2. Team flexibility
"We use multiple different systems and had limited capacity in our roadmap to work on a full integration, so we were looking for a team that was scrappy and willing to work with us to make it work."


‍3. Collaboration from day one
Despite some initial concerns about working with a young company, those worries quickly dissipated. "This concern went away pretty quickly as I tested the first functional prototype in under 2 weeks.”

{raul-estrada}

The Impact

{stat-cards}

The results exceeded Octave's expectations across multiple dimensions, and the transformation went far beyond metrics. Octave saw two major operational improvements:

{callout-cards}

But what impressed Raul most was Hello Patient's commitment to continuous improvement. "The thing that surprised me the most since going live is how much effort the HP team puts into continuous improvement," he shares. "The version of Mia we launched with (which was already pretty good) looks nothing like the version that is live today. The team has invested a lot of time on auditing the conversations and making improvements."

‍Impact on Revenue and Patient Experience
The deployment has delivered value across three key areas:


• Patient Experience: 

  Faster response times and immediate answers to questions

• Operational Efficiency: 

  Staff can focus on complex cases while Mia handles initial inquiries

• Business Growth: 

  Increased acquisitions from phone inquiries


"The collaboration didn’t stop once we launched. It has actually gotten stronger."

What's next?

Four months in, Octave continues to see improvements in Mia's capabilities and performance. The AI Assistant has become an integral part of their client onboarding process, creating a new channel for engagement that didn't exist before.

"We set out to find a tool that could respond quickly to client inquiries while maintaining the high bar of quality that we pride ourselves on. Only four months into our partnership with Hello Patient, I can confidently report that these expectations have not only been met, but exceeded."

When asked if he would recommend Hello Patient to others, Raul's answer was unequivocal:

{raul-estrada-2}

For mental health providers and other healthcare organizations looking to better serve their clients while improving operational efficiency, Octave's success with Hello Patient offers a compelling blueprint. Sometimes the best solutions come from teams willing to collaborate, iterate, and put client experience first—even if that means answering calls at 2 AM.

Frequently Asked Questions

What does an AI voice assistant do for a mental health practice's inquiry line?

An AI voice assistant answers the calls a mental health practice gets from people looking for therapy, so no inquiry sits in voicemail. Hello Patient builds AI agents that handle these conversations by voice, text, and web chat. On an inquiry line, the agent picks up the call, answers questions about services, availability, and how to get started, and points the prospective client toward the next step.

The work is mostly the front door of care. Someone calls to ask what the practice offers, whether it fits what they need, and how soon they can begin. The agent handles that conversation in a natural voice, with no hold time and no busy signal. When a call needs a real person, the agent hands off to your team with the full context of what was already said, so the client does not have to start over.

How does an AI voice assistant answer prospective-client inquiries in the evenings and on weekends?

It answers every inquiry call the moment it comes in, 24/7, including the evenings and weekends when a practice's phones would otherwise go quiet. Hello Patient's AI voice agent does not keep office hours. A prospective client who calls late on a weekend evening reaches the same natural-sounding voice and the same clear answers they would get in the middle of a weekday.

That timing matters more in mental health than in almost any other line of care. Someone who works up the nerve to call about therapy after work may not call a second time if they hit a voicemail. The agent takes that call, answers what the person asked, and helps them move forward while the intent is still there. Your team starts the next morning with those conversations already handled instead of a stack of messages to chase.

How does the agent handle a sensitive or urgent inquiry from someone in a hard moment?

The agent stays inside a clear scope. It answers questions about the practice's services, availability, getting started, and scheduling. Hello Patient builds it to never give clinical or medical advice and never act as therapy. If a caller raises something clinical, something urgent, or a crisis, the agent routes that person to a real member of your team, following the protocols the practice sets.

This is the line we hold most carefully on a mental health inquiry line. The agent's job is to make the first step easier for someone reaching out, not to counsel them. So it handles the practical questions well, and it hands off anything that belongs with a person, with the context of the call passed along so nobody has to repeat a difficult moment. The practice decides what counts as a handoff, and the agent works to that.

How does the AI voice agent answer questions accurately about a specific practice?

Every answer comes from the practice's own knowledge base, so what the agent says matches what the practice actually offers. Hello Patient loads that material in, and the agent draws on it to answer questions about services, availability, and how care works. A prospective client gets a real answer to a real question instead of a menu tree or a callback promise.

Two things keep the experience high. The voice is low-latency and natural, so the conversation feels like talking to a person rather than waiting on a system. And the answers come from the practice's own information, so they stay correct as the practice updates them. When a question falls outside what the agent knows or should answer, it hands the call to your team rather than guessing. That is how the client experience stays as good on the phone as it is in person.

Does an AI voice assistant work across the different systems a practice already runs, without a big integration project?

It does, and it does not require a heavy integration project to get there. Many mental health practices run across several systems and platforms at once, and Hello Patient's AI voice agent is built to work across that setup rather than around it. For Octave, a network of professional therapists serving thousands of clients, Hello Patient stood up a working agent in under two weeks, across Octave's multiple systems.

The reason it moves that fast is the approach. Hello Patient starts the agent on the inquiry line where the need is clearest, works across the systems already in place, and gets a functional version answering calls early. You do not wait on a long build before the agent handles its first real conversation. From there the practice tunes and expands it, with your team seeing how it works on live calls the whole way through.

How does an AI voice agent add inquiry-call capacity and a new text channel without adding staff?

It answers far more inquiry calls than a team could on its own, and it opens a text channel a practice did not have before, with no new hires. Hello Patient's AI voice agent takes the inbound calls a busy line would drop or send to voicemail, so more people who reach out actually get answered. Octave nearly doubled its capacity to answer inquiry calls this way, and saw more acquisitions come from phone inquiries as a result.

The text channel is the second gain. Alongside the calls, the agent runs an automated text conversation, a new way for prospective clients to engage that no one on the team has to staff. For Octave, that new channel reached a 50% engagement rate. Four months in, the program is still improving. For a practice, that means more of the people looking for care get a real response, on more channels, without pulling anyone off other work.

Is an AI voice assistant HIPAA-compliant for mental health information?

An AI voice assistant handling mental health inquiries has to protect sensitive information, and Hello Patient meets that bar. Hello Patient is HIPAA-compliant, is SOC 2 Type 2 certified, and signs a BAA on every account. That floor holds whether the agent is answering a first inquiry call or running a text conversation with a prospective client.

Mental health information carries real weight, so the compliance is not an afterthought. The agent works within the scope the practice sets, keeps to answering services and scheduling questions, and hands off anything clinical to a real person. Every interaction is handled under the same protections a practice already expects for protected health information. If your compliance team wants to walk through how the agent handles data, Hello Patient will go through it with you before anything goes live.