Panos Stravopodis

Making voice AI more human @elyos_ai

Co-founder & CTO Elyos AI, YCS23

[Panos Stravopodis @pstrav]

4 days to go and I'm really excited to announce the panel for the @Elyos_AI x @googlecloud Voice AI event on Tuesday. And what a lineup it is!
🎙️ @SaarasMehan - Co-founder & CTO @Jackandjillai
🎙️ Pieris Christofi - Forward Deployed Engineer @ElevenLabs
🎙️ Fabian Seipel- Co-founder & CEO @ai_coustics
🎙️ @dimiwonders - Co-founder & CEO @wondercraft_ai
🎙️ Mikel Bober-Irizar - Co-founder & CTO @iconicgamesio

I'll be keeping the panel honest, digging into what's actually working in production and what's next for the stack.

If you're building in the space, this is the room to be in. 👇
https://t.co/n2Y3EtlxVO

[Panos Stravopodis @pstrav]

@kwindla Exciting - congrats team!

[Panos Stravopodis @pstrav]

@Elyos_AI and @googlecloud are bringing London's Voice AI community together on April 28th.
This is for founders and builders: Real talk about what's working in production and what's next for the stack.
See you there 👇

https://t.co/r1hIfyxli7

[Panos Stravopodis @pstrav]

The hardest problems in Voice AI can't be solved by Claude. (We checked.) The @Elyos_AI team will be in San Francisco next Monday hosting a dinner for Founders and Founding Engineers deep in this space. We've got an amazing group coming to talk about what's actually hard in voice right now.
Come join us 👇 https://t.co/iMYJ0hi9mF

[Panos Stravopodis @pstrav]

@garrytan Love it! What about a /fundraise skill @garrytan? Might need it soon.

[Paul Graham @paulg]

It's Y Combinator's 21st birthday today.

[Panos Stravopodis @pstrav]

If you geek out on real-time audio pipelines, follow along.
Or even better: come build this with us 👇https://t.co/Xky6xCsdnA

[Panos Stravopodis @pstrav]

"It doesn't meet our bar."
That's what we said listening back to our early calls. The agent stepping on the user, missing cues, killing the flow.
We were determined not to ship that. 🧵

[Panos Stravopodis @pstrav]

This screenshot is a reminder of how far we've come.
Over the next few weeks I'll be sharing the models we tested, the latency tradeoffs, and what actually moved the needle in production.
The stuff that doesn't make it into the docs.

[Panos Stravopodis @pstrav]

@sukh_saroy Would be interesting to see how maintainable would that be.

[Panos Stravopodis @pstrav]

The @Elyos_AI team will be in San Francisco next month and we are hosting a dinner for Founders and Founding Engineers building voice agents.

Just good food, sharp minds and honest conversation about what’s actually hard right now in voice.

https://t.co/USpJnppIX0

[Panos Stravopodis @pstrav]

Don't miss out on this. This is a pivotal moment for founders to build something great. Deadline is February 9th. Apply now @ycombinator

[Y Combinator @ycombinator]

The way startups are built has shifted quickly.

We're excited about a range of startup ideas for AI-native companies that can now be built faster, cheaper, and with more ambition than ever.

https://t.co/Ms3Vn5Vq7q

[Panos Stravopodis @pstrav]

WER by Turn Length

[Panos Stravopodis @pstrav]

At @Elyos_AI We benchmarked 13 STT providers on 100 real customer calls from the trades businesses.

Not synthetic lab data. Real calls with:
- Background noise & multiple speakers
- UK postcodes & addresses
- Regional accents (England, Scotland, Ireland)
- Short confirmations to long explanations

Top performers:
🥇 @DeepgramAI Flux - 15.9% WER
🥈 @soniox_ai - 16.9% WER
🥉 @Speechmatics - 17.7% WER

@OpenAI Whisper? 39.8% WER - wouldn't recommend for production voice AI.

What's your experience with STT models? Are we there yet?

[Panos Stravopodis @pstrav]

WER by Turn Type (e.g postcode, address, name etc)

[Panos Stravopodis @pstrav]

@hasantoxr @hasantoxr have you seen any data from a production deployment? Anyone using it at scale?

[Panos Stravopodis @pstrav]

In case you missed it, 2026 started with a bang for us

[Elyos AI @Elyos_AI]

We’ve just raised a $13m Series A for Elyos AI!

We’re also hiring across the board! See here for the open roles https://t.co/4OLyjoUzjb
@AdrianJohnst @pstrav @pipclairebrown

[Panos Stravopodis @pstrav]

Crazy that you can actually run this locally. Will be sharing results soon.

[Hugging Models @HuggingModels]

NVIDIA just dropped PersonaPlex-7B 🤯

A full-duplex voice model that listens and talks at the same time.
No pauses. No turn-taking. Real conversation.

100% open source. Free.
Voice AI just leveled up.

https://t.co/YfzFQfBzMS