kwindla ✓

@kwindla · San Francisco, CA · joined 20 Sep 2008

Infrastructure and developer tools for real-time voice, video, and AI. @trydaily // ᓚᘏᗢ // @pipecat_ai

15 731Followers

3 924Following

6 599Posts total

377.1KViews on collected posts

Against accounts of the same size

13 posts from the last 90 days, next to the 10K–100K follower range. shown widely, but few of those viewers react.

Median views5 148this account1 540median for 10K–100K

Reach, %32.73%this account4.74%median for 10K–100K

Engagement, %0.91%this account1.87%median for 10K–100K

Metric	This account	Median for 10K–100K	Ratio
Median views per post	5 148	1 540	3.34×
Reach (views ÷ followers)	32.73%	4.74%	6.90×
Engagement rate	0.91%	1.87%	0.48×

Others in this range → Compare with another account → How these benchmarks are built →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

325.3K27 Aug

7.1K

10.2K

17730 Aug

80031 Aug

7.6K

966

6.8K

221 Sep

3.7K

5.1K2 Sep

5.3K

Last 13 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.90%27 Aug

0.59%

1.16%

0.85%

7.91%30 Aug

0.75%31 Aug

2.07%

0.52%

0.69%

0.00%1 Sep

2.07%

1.24%2 Sep

2.63%

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.87%.

What the audience does

Likes42.0%3 147 in total

Reposts3.9%294 in total

Replies2.0%153 in total

Quotes0.7%53 in total

Bookmarks51.4%3 853 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

Followers by day

4 Sep

Daily snapshots since 04 Sep 2026; the dashed line is the starting count.

Latest posts

Here's a complete voice agent tutorial focusing on a super low latency configuration: Pipecat PhoneLLM Alpha 1 on @modal, Deepgram transcription, Cartesia voice. You can grab the code and jus run /setup in Claude or Codex. But Merve's video also dives into the details of https: 5.3K views · 116 likes · 14 reposts · 9 replies 02 Sep 2026 If you're interested in voice UI, or music, or both, it's worth watching all of this video that Jon just dropped. Voice-controlled, generative music. The interface will be familiar to anyone who has used a digital audio workstation. But also ... deep integration of voice makes 5.1K views · 61 likes · 2 reposts · 1 replies 02 Sep 2026 I joined the Pipecat TV crew to talk about PhoneLLM: a super low-latency LLM trained on voice agent scenarios. We talked about building agents with a small-ish LLM like this, compared to using a big model. PhoneLLM is based on Nemotron 3 Nano, so it's a 30B-A3B MoE. Prompting a 3.7K views · 65 likes · 4 reposts · 8 replies 01 Sep 2026 @kwindla Latency is the number everyone measures, but on real phone lines the thing that breaks trust is interruption handling. A caller talks over the greeting, the agent keeps going, and the call is lost before the model even matters. Curious whether PhoneLLM helps with barge-i 22 views · 0 likes · 0 reposts · 0 replies 01 Sep 2026 Over a year ago, I posted a video where Gemini and I collaborated on an Ableton Live Session. Music AI is fun! Creating voice-native music software is something I'm really passionate about, so, here is Jamcat. Jamcat's concept is a voice controlled, generative music performance 6.8K views · 36 likes · 5 reposts · 4 replies 31 Aug 2026 PhoneLLM: https://t.co/fiI1HMj8EK Weights on @huggingface and one-click endpoint deployment on @modal. 966 views · 5 likes · 0 reposts · 0 replies 31 Aug 2026 Everybody building voice agents obsesses about latency. Every enterprise I talk to wants to run LLMs on their own infrastructure (for data management, regulatory/compliance, and cost reasons). PhoneLLM is a new open weights model trained on voice agent tasks: sub-100ms https:// 7.6K views · 138 likes · 10 reposts · 8 replies 31 Aug 2026 Full episode, including a fantastic segment with @AndyMasley about the data center build-out backlash, here: https://t.co/pc0cm5Gjo8 800 views · 5 likes · 0 reposts · 1 replies 31 Aug 2026 @kwindla yes impressive, voice agent, we are also heading there 🚀 177 views · 9 likes · 3 reposts · 2 replies 30 Aug 2026 Blog post with more details about this model is here: https://t.co/HnFhvvzdI3 We also have a new internal benchmark, PhoneBench, that's part of this project. And we're working with partners to create custom models using the "factory" that produced PhoneBench. If you need a http 4K views · 32 likes · 0 reposts · 2 replies 27 Aug 2026 Weights are here: https://t.co/y727Voi1bt 10.2K views · 113 likes · 3 reposts · 2 replies 27 Aug 2026 Ben Shababo at @modal did a bunch of great inference optimization work to achieve the >80 concurrent clients at sub-600ms end-to-end TTFAT. You can spin the model up as a Modal Endpoint with one click or with this command: ``` modal endpoint create --model https://t.co/XkOB 7.1K views · 39 likes · 2 reposts · 1 replies 27 Aug 2026 Introducing PhoneLLM, an open model for voice agents. GPT 5.6 Terra performance on typical voice agent tasks at 1/3 the latency and 1/18 the cost. For voice agents, we need models that are both very low latency and very good at tool calling and instruction following. There's a 325.3K views · 2.5K likes · 251 reposts · 115 replies 27 Aug 2026