Pierson Marks

@jellypodai founder & CEO | go bruins @ucla

[Pierson Marks @piersonmarks]

Introducing Jellypod.... The AI Podcast Studio.

From the comfort of your home, Jellypod empowers anyone to effortlessly produce professional quality podcasts and share their stories with the world. What story will you tell?

[Pierson Marks @piersonmarks]

@ajs6888 @ClaudeDevs exactly! the jellypod agent

[Pierson Marks @piersonmarks]

While everyone's talking about prompting agent loops for engineering, I tried it for GTM.

I set up a daily automation two months ago that tracked every single Claude Code release and automatically published a 3 minute podcast to Apple and Spotify on the top new features and bug fixes.

Today, we crossed 2,000 listeners on The Claude Code Changelog podcast.

Fully on autopilot using @JellypodAI automations.

[Pierson Marks @piersonmarks]

@rohitdotmittal I really hope some of these new deca millionaires pool together to start building - and going after some of the major issues in the city. Opportunity to growth their wealth and apply in new domains like real estate development or physical infrastructure

[Pierson Marks @piersonmarks]

@BhosalePratim @neilzegh @GradiumAI Absolutely

[Pratim🥑 @BhosalePratim]

You can now use @GradiumAI TTS via the Speech SDK🚀

[Pierson Marks @piersonmarks]

@haydenbleasel 🔥

[Pierson Marks @piersonmarks]

@vibemarketersHQ & repurposed into micro AI podcasts...

we generate literally thousands of inbound with niche content repurposed with @JellypodAI

[Pierson Marks @piersonmarks]

@jackfriks @postbridge_ How’d you double growth to $40k?

[Pierson Marks @piersonmarks]

@mattshumer_ Will need a head to head comparison to @JellypodAI Magic video

[Pierson Marks @piersonmarks]

@ScottChoppin Go bruins!

[Pierson Marks @piersonmarks]

@BhosalePratim @neilzegh @GradiumAI @pipecat_ai @deepwhitman 🎉🔥 I feel like a collab launch announcement is in order

[Pierson Marks @piersonmarks]

@neilzegh @GradiumAI @pipecat_ai @deepwhitman Would an engineer from your team be down to open a PR?

https://t.co/vAZdf4z0SD

[Pierson Marks @piersonmarks]

@thingsinmotion @infillworks This is so cool. We need to give Lennar and Toll Brothers a run for their money on the bland mass produced track home era

[Pierson Marks @piersonmarks]

@nandwani_janhvi @MiniMax_AI @cartesia Drop a follow and rt if you agree

[Pierson Marks @piersonmarks]

SpeechSDK v0.15 just dropped.

- New models: @MiniMax_AI speech-2.8, @cartesia sonic-3.5
- Pitch-preserving speech control (0.75 - 1.5x)
- Peak aware volume adjustment
- Faster long-form synthesis via parallel chunking
- Native per-provider sample rates (up to 48kHz)

[Pierson Marks @piersonmarks]

Open-source, details below:

https://t.co/boJFi8HlGV

[Pierson Marks @piersonmarks]

Text to Speech looks easy, until you need multiple models from multiple providers - all with different capabilities and API shapes.

@SpeechbaseAI takes everything we’ve learned building @JellypodAI to create the best DX working with generative audio.

Public beta is live!

[Bilal @deepwhitman]

Sneak Peak at what we've been cooking...

Tired of installing a different SDK for every Text-to-Speech provider?

Speechbase is the production stack for AI audio: one API across every major TTS provider (@OpenAI, @ElevenLabs, @cartesia, @DeepgramAI, @xai, @fal and more).

Swap providers with a string change. Word-level timestamps, multi-speaker, pronunciations, moderation, request logs, and analytics all built in. See every request, across every provider, in one place.

One API. Every voice. No lock-in.

[Pierson Marks @piersonmarks]

The SpeechSDK is fully open-source, supports every major TTS model provider, and helps you build better generative audio applications.

https://t.co/vAZdf4z0SD

[Pierson Marks @piersonmarks]

@cartesia's latest flagship model, Sonic-3.5, is now available on the SpeechSDK.

Sonic-3.5 has lead the Artificial Analysis's TTS leaderboards and supports 40+ languages, with improved naturalness, pacing, and emotional expression.

[Pierson Marks @piersonmarks]

To use Sonic 3.5, just swap a single model string in the SpeechSDK and choose your voice.

Or mix-and-match providers to create multi-speaker conversations with the generateConversation() API.