Building the future of voice AI.
We are a team of engineers and researchers dedicated to making phone AI smarter, faster, and more accessible for everyone.
Our Story
We’re a team of Stanford AI Lab and Y Combinator alumni that have been working on voice AI since 2024. Originally, our focus was on Flux Interpreting, which translates phone calls between languages.
But after being repeatedly asked by customers to automate phone calls for their businesses instead of just translating them, we shifted focus to what is now ThunderPhone.
Initially, we began by building phone agents for customers on other company’s phone AI platforms. We tested just about every phone AI platform on the market, but unfortunately found them lacking. The limitations of single transcription models, single LLMs, and difficult integrations meant we couldn't deliver the quality our clients needed.
So, we built our own custom stack. And we realized: existing platforms were like asking web startups in the 90s to host their own servers. It was time for a platform that just works.
The Integrated Advantage
Other phone AI platforms limit the performance of their systems by enforcing a rigid 3-step structure: One transcription model → One LLM → One TTS model.
We’ve taken a different approach. In line with Clay Christensen’s modularity theory, we believe it’s still early days in voice AI, and it’s more valuable to maximize performance with an integrated architecture than to maximize customizability with a modular one.
Our system uses many models continuously:
- Combining multiple transcription models to balance speed and accuracy.
- Combining audio-input LLMs (better hearing) with text-only LLMs (smarter reasoning).
- Combining fast models for quick replies with slower, smarter models for oversight.
- Using specialized models to classify and filter background noise.
This tightly-integrated architecture gives our system far higher performance at lower cost than the rigid approach offered by others.
Optimized Decisions
As far as customizability goes, we are against it! Most of today’s phone AI platforms force users to navigate endless choices, from the exact models used to ultra-detailed parameters.
In our opinion, it is our job to optimize these choices behind the scenes. We see ourselves as a sort of Squarespace for voice AI—making it easy to set up phone agents at the absolute highest performance without lifting a finger.
We also believe in competitive pricing. All of our service levels are priced at between 10 and 25 percent gross margin. This means that even if you rebuilt ThunderPhone from scratch, you’d effectively be paying the same amount, without the benefit of our behind-the-scenes optimizations.
Ease of Use
ThunderPhone is incredibly easy to use, whether you’re setting up a new phone agent or migrating from another platform.
All you need to do is:
- Decide whether you want Zap, Spark, or Bolt.
- Choose a voice.
- Give us your prompt.
- Connect software integrations and VoIP with our easy wizards.
All of this typically takes 10 minutes or less.