Sonia is a conversational AI for mental health support. Our ambition is to build an AI-based system that is orders of magnitude more effective than any existing digital health intervention.
External link for Sonia (YC W24)
186 N 6th St
Brooklyn, New York 11211, US
Sonia (YC W24) reposted this
Why Pay for Billboards? We Went Full DIY with 5 U-Hauls in NYC $49.99 for a mobile ad campaign, more impressions, and way more fun! And the best part - a month of Sonia costs even less! Download Sonia (YC W24) https://lnkd.in/ezSBqvMG
Sonia (YC W24) reposted this
We are hiring... ... a person who wants to help us with part 3 of the Sonia (YC W24) U-Haul campaign 🎊 We need someone who helps us drive one of our trucks in Manhattan on Friday and Saturday nights. I'll be driving one of the trucks as well :) Insurance and everything else is covered. We'll have a nice dinner together afterward. DMs are open!
Sonia (YC W24) reposted this
In the U.S., over $1 trillion is spent on healthcare admin annually, with clinicians spending up to 50% of their day on documentation instead of patient care. AI has the power to automate nearly half of these tasks, saving billions. 💡 Innovators like Chartnote, Sonia (YC W24), and StackAI are taking the healthcare industry by storm with AI-driven medical scribes, QA tools, and healthcare agents, helping clinicians get back to what they do best—caring for patients. And we’re fueling this innovation. Our Nova-2 Medical model is purpose-built for real-time healthcare transcription, capturing everything from symptoms to treatments with unmatched precision and speed. Learn why companies rely on Nova-2 Medical to power their healthcare applications. https://lnkd.in/gGqj4z7S #medteach #healthcareinnovation
Here's a concise, high-level description of how our therapists work under the hood!
Don’t spend a **** of money on generic voice AI providers. Build it yourself. What are the basic components? At the heart of a voice AI stack is an LLM that engages in the conversation by generating responses. Whether your LLM is a custom RAG or some other fancy architecture is up to you. LLMs operate on text, so you need a second model that transcribes speech into text. There’s a tradeoff between quality and speed, but if you want to build a real-time voice application these days, I don’t see anything else being used than Deepgram. What’s next? You need to take your generated text and synthesize a voice — unless you want to read the responses aloud to your clients. This is NOT what Paul Graham means by saying, “Do things that don’t scale.” There are amazing voice APIs, e.g., ElevenLabs high-quality voices and recently launched Cartesia with BLAZINGLY fast voice generation. But how do you move audio between the client and AI/server? WebRTC is the standard for sending real-time data online. Daily has an amazing WebRTC platform with client libraries that abstract playing audio for you (try that in Swift). So, what’s missing? We covered HOW to generate a response, but you must also decide WHEN to respond. Use a voice activity detection model like Silero that tells if your client is speaking. During silence, you can query a small, fast LLM to evaluate semantically if the client really finished speaking, and your AI should respond. Turn-taking is very domain-specific. At Sonia (YC W24), we’re building an AI therapist, and pace and turn-taking fundamentally differ from customer support or sales agents. To make turn-taking work really well, you want to incorporate the full audio signal, as the spoken word contains important information beyond just the text. Depending on your domain, a custom-built turn-taking model might make sense. Care about latency. Perplexity CEO Aravind Srinivas cares A LOT about latency. Improve latency by hosting models yourself. Baseten has a simple interface for model hosting with a large library of open-source models. But will GPT-4o voice mode come and haunt us? Embrace the progress in the field. It will be hard to beat multimodal model latency as long as you use a similarly complex LLM in your custom stack. But I don’t think latency matters since you get 1-2 second latency building it yourself. However, depending on the application, there will be huge benefits of feeding raw (or preprocessed) audio into the model without having text as the bottleneck modality. Many apps already go beyond voice with video avatars like Tavus. You can expect increased engagement for many applications (e.g. healthcare, education). The space is quite hot right now. Our YC friends at Arini (YC W24) handle phone calls for dentists, and Scritch (YC W24) recently launched AI assistants for vets. My calendar is open if you want to connect or get advice!
I’m struggling with social situations. What if you could just send an AI agent to meet your future in-laws? Here's what it would look like. 🏠 The Dinner It’s the first time you've met Sam, your partner's father. Sam is the CEO of TokenAI, one of the larger AI companies in your area. Since you spent your college years studying Machine Learning instead of meeting potential in-laws, you built an AI voice agent to help you tonight. 👂Listening At the dinner table, Sam is rambling about the new multimodal model they're about to launch. You've done your research and know that Sam is not a man of patience - so you designed your AI system to be blazingly fast. You connected your smartphone to a streamed transcription model that continuously listens and sends you Sam's transcribed speech. A few minutes in, Sam seems to be taking a break while forking a bite of his plant-based protein bowl, which he got from TokenAI's employee-benefit health food station. Intimidated by the authority and unsure when to speak up, you're running a voice activity detection model on your phone, which just at this moment signals you it's time! 🧠 Cognition While Sam's grinding his spoon in the protein bowl (wait... did you just spot a cricket leg in there?), you send the transcribed text to the remote LLM system you prepared before dinner. Your LLM is based on a custom parents-in-law retrieval model fine-tuned on all three seasons of MTV’s "Date My Mom” show, allowing the LLM to deliver state-of-the-art reasoning capabilities in this domain. You were smart enough to host your model with a fast cloud-based model inference provider, so you see the first response tokens being streamed to your phone after just the blink of an eye. 🗣️ Speaking Sam seems distracted and furiously fidgets with an acai berry in his bowl. “Why is my dearest child dating this idiot?”, he thinks. You have to be fast now; don't ruin your first impression! While your LLM is streaming the text, you instantly send them to a cloud-based voice generator that sends you the audio chunks you can play to Sam. In preparation for this important dinner, you skipped today's yoga class and instead cloned your voice. Playing the synthetic voice, you realize you generated an intriguing, realistic-sounding response. While playing it from your phone's speakers, you struggle to keep your lips in synch, but Sam, checking the TokenAI Slack channel while listening, doesn’t notice. A minute into your speech, he starts coughing wildly before cutting you off (which your voice activity detection model noticed even before you). Sam is incredibly impressed by your brilliant response and offers you a leading position at TokenAI. You graciously acknowledge the offer but politely decline, explaining that you're fully committed to working on your latest startup idea, "AirBnb for Goldfish Enthusiasts". ⏭️ Jokes Aside Don't miss part 2 of this post - I will detail how we built our voice AI at Sonia (YC W24).
Sonia (YC W24) reposted this
We'll pay 1% of your rent in NYC... All you need to do is hang up a Sonia “artwork” outside of your apartment. I’ll personally deliver the artwork and bring you your favorite coffee. We have all sorts of different dimensions and slogans (“listens better than your ex” is my favorite). You’ll obviously also get a free Sonia subscription. Reduced rent + free therapy sessions is the ultimate mental health combo we all need living in this city. Side benefit: your neighbors might become less anxious after using Sonia and actually start saying hello when you walk past them :) Only available as long as supply (aka funding) lasts!
It’s here. Sonia 2.0 has arrived. 🚀 Today marks a new era in accessible mental health support. We're thrilled to introduce the new and improved Sonia – a major leap forward in our mission to make professional-grade mental wellness tools available to everyone, anytime, anywhere. The Sonia team has poured our hearts into this update, and we could not be more excited to share it with you. Our goal is to help bring happiness and well-being to as many people as possible, at a scale that truly makes a difference. 💙 What’s New in Sonia 2.0? ✨ 🌍 World’s First AI-Assisted Program for Generalized Anxiety Disorder (GAD): We’re introducing the first AI-driven CBT program designed specifically for GAD. This series of six structured sessions is based on evidence-based treatments, offering you practical support to manage anxiety at your own pace. 🎨 Redesigned Interface: Sonia 2.0 features a completely overhauled design with a focus on user experience. Switching between chat and voice modes is now smoother, and navigating the app is more intuitive, making it easier to engage with your therapy sessions. 🎯 Personalized Onboarding: From your very first session, Sonia 2.0 adapts to your needs. Choose your focus areas, preferred therapeutic style, and session structure to ensure the support you receive is tailored specifically to you. 💬 Community Canvas: We’re introducing a new way to connect with others through our Community Canvas. Share your thoughts on the “Question of the Day” and engage with a supportive community focused on mental wellness. ⏱ Short Sessions for Busy Lives: We’ve added new session types—structured advice for problem-solving and open-ended sessions for emotional release. These are designed to fit into your day, providing support whenever you need it. 🔔 Enhanced Notifications & Scheduling: With Sonia 2.0, you can now set reminders that help you stay consistent with your therapy. Schedule sessions when it’s convenient, or jump into one whenever you’re ready—flexibility is built-in. 🎙 New Therapist Voices: We’ve expanded our range of therapist personas and voices, allowing you to choose the one that best suits your comfort level, making your sessions more personalized and engaging. ... among other improvements, but we'll leave that for you to explore 😊 This new version of Sonia is just the beginning. While we have a long journey ahead of us, we hope it marks a meaningful step toward greater happiness and well-being for all our users! Download Sonia 2.0 today on the App Store: https://lnkd.in/eUMvdp2M #MentalHealth #AI #Therapy #SoniaApp #CBT #HealthTech #Innovation
Something Big Has Arrived…✨ We’ve been hard at work behind the scenes, and we’re excited to finally share that Sonia has just taken a major leap forward. Introducing Sonia 2.0 – our most innovative update yet, designed to elevate your mental health journey even further. 🌟 With new features and enhanced tools, including something special for managing anxiety, Sonia is now more equipped than ever to be there for you. Curious? Stay tuned for the full reveal on Monday – trust us, you won’t want to miss this! Full Details Coming Soon… 💙 #MentalHealthMatters #AI #TalkToSonia #ComingSoon
On a scale of 1-100, what percent happy are you right now? In today’s largely data-driven world, we’ve unsurprisingly managed to turn happiness into yet another metric to optimize. Apps track our moods, algorithms suggest how to boost our spirits, and we're constantly urged to maximize our "happiness score." But as Jessica Grose questions in her recent article “Are We Happy Yet?” published in the New York Times: Is this obsession truly making us happier, or could it actually be making us feel worse? The pursuit of perpetual positivity might be creating a new kind of anxiety – the stress of not feeling happy enough. In reality, true mental health isn't about reaching a happiness high score. As Grose puts it, “Feelings aren’t the same as other kinds of health metrics, like steps and heart rate and liver function…the user experience of being alive cannot be graphed.” Perhaps the path to genuine well-being lies not in relentless self-optimization, but in fostering self-acceptance and human connection. As so many of us attempt to navigate and make sense of this era of digital emotion, let’s consider: What if the key to a fulfilling life isn't about maximizing happiness, but about finding meaning in our full range of experiences? At Sonia, we're inspired by these insights to rethink how technology can support mental health. Our goal isn't to graph your emotions or optimize your happiness score, but to provide accessible, empathetic support that honors the full complexity of the human experience. Because at the end of the day, in our pursuit of the perfect emotional state, we might just be missing the beautiful intricacy of simply being human. Read the full article by Grose here: https://lnkd.in/dRb4C_Kq #mentalhealth #happiness #technology
If your idea of “finding your center” involves locating the least chaotic corner of your living room, you’re not alone. We've compiled 6 mindfulness techniques for those of us who can't touch our toes or sit still for more than 30 seconds. Highlights include: - Breathing: You're doing it anyway, might as well pay attention - Thought observation: Procrastinate on your own anxiety - Five senses check-in: Remind yourself you're not actually living in a simulation - Body scan: Like a TSA pat-down, but for your own tension - Gratitude moment: 60 seconds to appreciate life before returning to your existential crisis - Mindful eating: Pretend you’re a food critic who gets paid by the adjective, minus the pretentious vocabulary No lifestyle overhaul required. Just practical ways to avoid that mental breakdown you've been scheduling for next Tuesday. 💙 Read our new blog article here: https://lnkd.in/eyTnVaqW