AI therapy you can actually talk to. Just speak naturally and get support anytime you need it.
Replies
Best
I am very happy and proud to be one of the first users. Lovon helped me a lot to save my relationship when my girlfriend and I were at a distance. And now we're getting married soon!
@leonid_ivanov Great question! We actually see Lovon as the bridge to human therapists, not a replacement. Right now we focus on being there 24/7 when professionals aren't available, plus our crisis detection automatically connects people to live help when needed.
Would you see live chat as an escalation for specific situations, or more as an ongoing option?
@leonid_ivanov It's interesting idea! We will need to think about how to have ready to dive into therapists and how to give them context, but I think it should be as you said!
@nikvoice Haha right?! Mental health feels like one of the most human applications of AI - using it to actually understand emotions, not just write code.
Have you tried any AI therapy tools before, or would this be your first?
Checked the video and your website, looks like a very good solution to me, wish you success!
But the reviews on your site need links to the sources!
(also the 3 dots do nothing, and it's not clear what the orange checkmarks and "15 min with therapist" mean)
Also, I couldn't find real info about the "PhD psychologists" it is "built with", there need to be links to their professional profiles and scientific work for this statement to have credibility!
@konrad_sx Thanks for the detailed feedback - super valuable! 🙏
You're absolutely right about the credibility piece. We're actually building out a new "Our Team" section on the site right now where we'll feature our clinical experts with their full profiles and credentials. Plus, we're currently writing our first scientific papers to validate our approach.
1) Can the app recognise patterns of more serious mental disorders such as eating disorders, personality disorders, depression etc.? If yes, can it forward or give any contact/reference to a psychiatrist (for example)? I know that in many cases people do not know yet what mental condition they might have and usually general psychologists become the first point of contact so to say. But then they should be directed to appropriate professional help
2) Do you plan to expand the app to the desktop version as well?
3) Do you plan to allow texting besides voice-only interactions?
Thank you in advance for the responses! And congratulations on launching again, the app looks very polished and well-thought!
Lovon is a wellness tool, not a medical diagnostic device. That said, Anna (our AI therapist) is trained to recognize when someone might benefit from professional support - like if conversations reveal ongoing struggles. In those cases, we gently suggest connecting with a licensed therapist or other resources.
We actually have a working desktop version at app.lovon.app! It's MVP with limited features right now - you'll need to register through mobile first, then you can use desktop for sessions. Full desktop experience is coming.
We tested text extensively in earlier versions, but users consistently told us they don't want to type about their problems. Talking out loud feels more natural and therapeutic - like venting to a friend vs writing an email. So we're doubling down on voice instead.
Which matters more to you personally - desktop access or text option?
Report
@ponikarovskii Hello Anton, thank you for the reply! It's great that your tool can gently suggest connecting with a licensed therapist or other resources. Personally, I don't mind using mobile version or voice-only app, but i was just thinking about certain groups of people (older people, disabled persons, etc.) who might find it easier to use desktop or text-based alternative :) But i agree with you that talking out loud feels more natural and therapeutic
Where do you draw the line on what Lovon should *not* try to handle (certain conditions, situations, user states), and how are you testing/evaluating the system to reduce the risk of confident-but-wrong guidance as you scale and run clinical validation?
@curiouskitty Hi! Thanks for the great and correct question!
We don’t replace a real licensed therapist: we don’t diagnose and we don’t suggest medication. For that, of course, you still need to see a therapist. Also, if a user starts talking about things we legally can’t work with (something harmful), we stop the conversation right away and direct them to the right services.
But overall, we’ve only had a few situations like that.
Great question. We analyze every conversation in real-time for three critical markers: potential harm to self or others, suicidal ideation, and signs of serious mental health conditions. When detected, we immediately tell users to seek professional help - we're very clear about not being licensed therapists.
For testing: we run automated sentiment analysis comparing start vs. end of sessions (declining sentiment flags a problem), plus our clinical advisors regularly review anonymized transcripts and give us feedback. We're actively expanding our clinical board to strengthen this oversight.
What's one thing you wish more AI mental health products would be transparent about upfront, that would make the PH community feel safer trying them?
Congrats team! Curious — what happens when Anna doesn't know how to respond? Does she acknowledge her limits or redirect?
Report
Maker
@artyom_zhuravlev Hey Artyom! You need to try real hard to get Ann in this situation :) But in case of possible danger or if she believes that human specialist needed Anna can support redirect!
@artyom_zhuravlev absolutely, if there're any signs of harm - it stops the conversation immediately
Report
Really compelling approach to mental health support. The voice-first interaction feels like the right call — much closer to how real therapy works, and I imagine it lowers the barrier to opening up.
As someone building AI for health detection in a different domain (veterinary — detecting cat pain through facial analysis), I'm fascinated by how you handle the balance between genuine empathy and surface-level agreement. The CBT framework sounds like it adds real therapeutic substance.
Question: how did working with a PhD psychologist change your product design assumptions? Did you find that what felt "right" as engineers was sometimes wrong therapeutically?
@go_sakioka Thanks so much! You nailed the key insight - I've been in therapy myself since 2021, so I understood what users need, but had no idea how therapists actually build therapy.
Working with our PhD psychologist was like getting a crash course - I basically became a junior psychologist through our countless calls. The biggest revelation was that therapeutic alliance (the human relationship) matters more than any technique.
Your cat pain detection work sounds fascinating - do you find similar challenges in balancing algorithmic confidence with the subtlety of emotional/physical states that resist clear categorization?
Report
Totally relate to the "therapeutic alliance" insight — in our case it's the bond between cat and owner that drives daily usage more than the AI accuracy itself. To answer your question: yes, exactly the same challenge. Cat facial expressions are super subtle, so we use a confidence score rather than a binary "pain/no pain." The gray zone is huge, and we learned that showing uncertainty honestly actually builds more trust with users than pretending to be certain.
Replies
I am very happy and proud to be one of the first users. Lovon helped me a lot to save my relationship when my girlfriend and I were at a distance. And now we're getting married soon!
Lovon AI therapy
@narek_meliksetyan Wow, this made my day! Congrats on the engagement 🎉 Long distance is so tough emotionally.
I'd love to know - was it the 24/7 availability that helped most, or was there something specific Anna said that gave you a breakthrough?
@ponikarovskii thank you!
I think that everything combined has played its part.
Lovon AI therapy
@ponikarovskii @narek_meliksetyan thank you Narek! You're our power-user :)
@narek_meliksetyan It's such a big pleasure to read that it really someone!
That’s a great idea! How do you feel about adding live people chat on user request?
Lovon AI therapy
@leonid_ivanov Great question! We actually see Lovon as the bridge to human therapists, not a replacement. Right now we focus on being there 24/7 when professionals aren't available, plus our crisis detection automatically connects people to live help when needed.
Would you see live chat as an escalation for specific situations, or more as an ongoing option?
Lovon AI therapy
@leonid_ivanov thanks Leonid!
@leonid_ivanov It's interesting idea! We will need to think about how to have ready to dive into therapists and how to give them context, but I think it should be as you said!
Starnus
Amazing! Congrats!
Lovon AI therapy
@khashayar_mansourizadeh1 Thank you! 💙
Have you had a chance to try voice therapy yet, or what interests you most about the AI therapy space?
Lovon AI therapy
@khashayar_mansourizadeh1 thank you, Khashayar!
MeMemes
Finally, AI product not about coding!!
Lovon AI therapy
@nikvoice Haha right?! Mental health feels like one of the most human applications of AI - using it to actually understand emotions, not just write code.
Have you tried any AI therapy tools before, or would this be your first?
Lovon AI therapy
@nikvoice hahahahah true!! Thanks :)
App Finder
Checked the video and your website, looks like a very good solution to me, wish you success!
But the reviews on your site need links to the sources!
(also the 3 dots do nothing, and it's not clear what the orange checkmarks and "15 min with therapist" mean)
Also, I couldn't find real info about the "PhD psychologists" it is "built with", there need to be links to their professional profiles and scientific work for this statement to have credibility!
Lovon AI therapy
@konrad_sx Thanks for the detailed feedback - super valuable! 🙏
You're absolutely right about the credibility piece. We're actually building out a new "Our Team" section on the site right now where we'll feature our clinical experts with their full profiles and credentials. Plus, we're currently writing our first scientific papers to validate our approach.
App Finder
@ponikarovskii That's great to hear!
Lovon AI therapy
@ponikarovskii @konrad_sx thank you!!
Very impressive product! congrats on the launch!
I have a couple of questions though:
1) Can the app recognise patterns of more serious mental disorders such as eating disorders, personality disorders, depression etc.? If yes, can it forward or give any contact/reference to a psychiatrist (for example)? I know that in many cases people do not know yet what mental condition they might have and usually general psychologists become the first point of contact so to say. But then they should be directed to appropriate professional help
2) Do you plan to expand the app to the desktop version as well?
3) Do you plan to allow texting besides voice-only interactions?
Thank you in advance for the responses! And congratulations on launching again, the app looks very polished and well-thought!
Lovon AI therapy
@ksenia_sh Thanks so much! Great questions:
Lovon is a wellness tool, not a medical diagnostic device. That said, Anna (our AI therapist) is trained to recognize when someone might benefit from professional support - like if conversations reveal ongoing struggles. In those cases, we gently suggest connecting with a licensed therapist or other resources.
We actually have a working desktop version at app.lovon.app! It's MVP with limited features right now - you'll need to register through mobile first, then you can use desktop for sessions. Full desktop experience is coming.
We tested text extensively in earlier versions, but users consistently told us they don't want to type about their problems. Talking out loud feels more natural and therapeutic - like venting to a friend vs writing an email. So we're doubling down on voice instead.
Which matters more to you personally - desktop access or text option?
@ponikarovskii Hello Anton, thank you for the reply! It's great that your tool can gently suggest connecting with a licensed therapist or other resources. Personally, I don't mind using mobile version or voice-only app, but i was just thinking about certain groups of people (older people, disabled persons, etc.) who might find it easier to use desktop or text-based alternative :) But i agree with you that talking out loud feels more natural and therapeutic
Lovon AI therapy
@ponikarovskii @ksenia_sh actually we're planning to even improve our web(desktop) version! So it will fits all groups of people!
Congratulations on the launch 🎉 🎉
Lovon AI therapy
@shubham_pratap thanks a lot!
Lovon AI therapy
@shubham_pratap thank you man! Appreciate it!
@shubham_pratap Thank you!
Product Hunt
Lovon AI therapy
@curiouskitty Hi! Thanks for the great and correct question!
We don’t replace a real licensed therapist: we don’t diagnose and we don’t suggest medication. For that, of course, you still need to see a therapist. Also, if a user starts talking about things we legally can’t work with (something harmful), we stop the conversation right away and direct them to the right services.
But overall, we’ve only had a few situations like that.
Lovon AI therapy
@curiouskitty Honored to have you here!
Great question. We analyze every conversation in real-time for three critical markers: potential harm to self or others, suicidal ideation, and signs of serious mental health conditions. When detected, we immediately tell users to seek professional help - we're very clear about not being licensed therapists.
For testing: we run automated sentiment analysis comparing start vs. end of sessions (declining sentiment flags a problem), plus our clinical advisors regularly review anonymized transcripts and give us feedback. We're actively expanding our clinical board to strengthen this oversight.
What's one thing you wish more AI mental health products would be transparent about upfront, that would make the PH community feel safer trying them?
Remy AI
Congrats team! Curious — what happens when Anna doesn't know how to respond? Does she acknowledge her limits or redirect?
Lovon AI therapy
@artyom_zhuravlev @aleksanadr_lavrinenko thank you Artyom!
Remy AI
@thxgrey I wish you success in developing the project. It definitely deserves attention.
Lovon AI therapy
@artyom_zhuravlev thank you Artyom!
Remy AI
@aleksanadr_lavrinenko Thank you!
Lovon AI therapy
@artyom_zhuravlev absolutely, if there're any signs of harm - it stops the conversation immediately
Really compelling approach to mental health support. The voice-first interaction feels like the right call — much closer to how real therapy works, and I imagine it lowers the barrier to opening up.
As someone building AI for health detection in a different domain (veterinary — detecting cat pain through facial analysis), I'm fascinated by how you handle the balance between genuine empathy and surface-level agreement. The CBT framework sounds like it adds real therapeutic substance.
Question: how did working with a PhD psychologist change your product design assumptions? Did you find that what felt "right" as engineers was sometimes wrong therapeutically?
Lovon AI therapy
@go_sakioka Thanks so much! You nailed the key insight - I've been in therapy myself since 2021, so I understood what users need, but had no idea how therapists actually build therapy.
Working with our PhD psychologist was like getting a crash course - I basically became a junior psychologist through our countless calls. The biggest revelation was that therapeutic alliance (the human relationship) matters more than any technique.
Your cat pain detection work sounds fascinating - do you find similar challenges in balancing algorithmic confidence with the subtlety of emotional/physical states that resist clear categorization?
Totally relate to the "therapeutic alliance" insight — in our case it's the bond between cat and owner that drives daily usage more than the AI accuracy itself. To answer your question: yes, exactly the same challenge. Cat facial expressions are super subtle, so we use a confidence score rather than a binary "pain/no pain." The gray zone is huge, and we learned that showing uncertainty honestly actually builds more trust with users than pretending to be certain.
Lovon AI therapy
@go_sakioka you're right guys! wise conversation!