Blobfish is a human-like AI Dialer that engages in back-and-forth conversations with your prospects. It can call thousands of numbers in minutes, filter out bad numbers or voicemails, and ask lead qualifying questions - 10x your junior SDR’s outbound work.
Talk to our AI Dialer Now 👉🏼 www.blobfish.ai
Subscribe
Sign in
Hi Product Hunt 👋
I'm excited to share the story behind Blobfish AI and why we built it.
At our last startup, US Air Supply, we scaled to $1M ARR in 18 months through cold calling. We ran a team of 13 SDRs in a boiler-room setup, grinding out hundreds of calls each day.
But the process was painfully inefficient.
Our connection rate was just 2% with numbers purchased from ZoomInfo and Apollo. SDRs spent most of their time calling voicemails, dead numbers, front desks, or the wrong contacts. Out of all the conversations, only 14% led to a booked meeting, and only 60% would show up. That meant high CAC and low morale.
We also tried VA, traditional dialers, and enrichment tools—but nothing really worked. Dialers had delays that killed conversations. Lead enrichment was slow and manual. On average, each SDR booked just 19 meetings a month. No matter what we did, cold outbound was a grind.
We felt the pain deeply—burnout and churn were constant. Asking reps to hit 400 dials and 100 talk minutes a day just wasn’t sustainable. It was hard to retain talents, and every new hire felt like starting from zero.
We knew there had to be a better way. So we built it.
Blobfish started as a simple tool to weed out dead numbers—but over time, it evolved into a powerful AI Dialer that sounds human-like, engages in real conversations, and qualifies leads at scale.
Blobfish doesn’t just call. It talks like your top sales rep, asks qualifying questions you can’t find in any database (like ERP vendors or Cloud providers), and books meetings directly onto your team’s calendar.
Our dream was simple: wake up to a calendar full of qualified meetings—without grinding through 1,000 calls.
That’s the future we’re building with Blobfish.
— Sam & Angel
Founders, Blobfish AI
We used to spend hours grinding through cold calls with low connection rates and even lower conversions. Blobfish AI flips the script — it's like having a superhuman SDR on your team 24/7. It actually talks to leads, qualifies them, and books meetings straight to your calendar. No more dead numbers or endless voicemails. Just real conversations and real results. Huge win for any sales team!
👏 Big shoutout to @samuel_chien & @angellin821 for building something this powerful.
inncivio
@anna_raafat Love this 🙌
@anna_raafat Thanks so much for supporting Anna 🙌🏼 In terms of language, we already have customers deploying Chinese, and we're planning to roll out Spanish, Vietnamese, German soon. Is there a language you'd love to see?
@anna_raafat yes Anna, we actually already have existing customer that call homeowner with Mandarin to confirm about home tour. We are actually also rolling new features for language support as well as using local phone number. This is a bit trickier because of the compliance.
inncivio
@anna_raafat Noted 🫡 Will incorporate your feedback and keep you posted on the launch timeline for these two.
@angellin821 @anna_raafat what's the main sceneario for French and Portuguese? Do you need to call people in France or calling people in US but speak in French or Portuguese?
inncivio
@samuel_chien mainly for customers living in Portugal and France
Scene
This looks really slick! The voice sounds really human-like, although a bit salesy to my taste. Do you have in your roadmap a way to select from different conversational styles, or an ability to customize this?
Also pricing looks pretty friendly, but I was also wondering, if there is any guarantee of the conversion rate (probably not as it depends on offer desirability which you can't control).
Last thing I'm wondering about is the response speed. Currently there are quite noticeable pauses before answers which is a LLM limitation you probably can't do anything about. But it would be nice if you can elaborate on when do you think this will improve.
Great work, and good luck with the launch! 🚀
@martinmaricak Hi Martin, thank you so much for the comment. We actually test few different voices as well as cloning, and we pick the highest performing voice.
We see at least 40% connection rate, which is a 20x boost from industry standard. Most of customers are able to hit 70% connection rate. Love to see what's the connection rate you would need to make this worth onboarding?
For the pause, we are currently doing around 1300ms. This can be improved with fine tuning. If this is something interesting, for the enterprise customer, we do train custom voice model for latency closer to 600ms. For anything faster, we really are limited by the model we pick.
Thank you again.
@martinmaricak Thanks for the thoughtful comment Martin. Appreciate all feedback 🙌🏼