# Sequen chose Rust for low-latency feature serving

Markdown copy of https://realalpha.app/s/sequen-chose-rust-for-low-latency-feature-serving/ on Real Alpha, the best minutes from long interviews with the people building AI. Cite and link the page, not this copy.

- Speaker: [Zoe Weil](https://realalpha.app/people/zoe-weil/) · Zoe Weil is co-founder and CEO of Sequen.
- Host: [Sam Charrington](https://realalpha.app/people/sam-charrington/) · Host, The TWIML AI Podcast
- Show: The TWIML AI Podcast
- Episode: [Optimizing AI Recommendations for Revenue with Sequen Founder Zoe Weil](https://realalpha.app/v/MAABmDB51LY/), Oct 15, 2026
- Moment: 18:53 to 21:42 of the episode (3 min), playing on https://realalpha.app/s/sequen-chose-rust-for-low-latency-feature-serving/; the episode from there: https://realalpha.app/v/MAABmDB51LY/?t=1133

## Quote

> “I genuinely don't. I don't think that we could have because we, you know, **ultra, ultra low latency** feature transformations serving are so core to what we do.”
> — [Zoe Weil](https://realalpha.app/people/zoe-weil/) at [19:45](https://realalpha.app/v/MAABmDB51LY/?t=1185) · The TWIML AI Podcast

## Takeaways

1. Low-latency feature transformations are central to the backend.
2. Zoe says Python could not meet the serving needs.
3. The team uses Python for R&D and Rust for its core backend.

## About this segment

Zoe Weil says Rust's low-latency support made Python impractical for Sequen's feature transformations, though the team still uses Python for R&D. The example ties a language choice to its serving constraint.

From “Optimizing AI Recommendations for Revenue with Sequen Founder Zoe Weil”, Oct 15, 2026. [Watch the full episode](https://realalpha.app/v/MAABmDB51LY/)

## Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

[18:58](https://realalpha.app/v/MAABmDB51LY/?t=1138) in trying to put this, you know, platform or environment in place? I think we got lucky in a sense because maybe, like, six, seven years ago. Rust didn't have a ton of support in, like, open source AI libraries like Hugging Face. That began to change rapidly in the past, I would say, three to four years. And, you know, this is why I always say whenever you make a successful startup, like, there's quite a bit of luck there as well.

[19:33](https://realalpha.app/v/MAABmDB51LY/?t=1173) Timing is everything. Yeah, timing is everything. We got very, very lucky because we started building this company at a time when support for Rust and many, many open source frameworks, AI, ML frameworks. You think you could not have done it in Python? No, I don't. I genuinely don't. I don't think that we could have because we, you know, ultra, ultra low latency feature transformations serving are so core to what we do.

[20:03](https://realalpha.app/v/MAABmDB51LY/?t=1203) I do not think we could have done it in Python. Certainly, we do a lot of R&D in Python, but the core of our backend is all in Rust. So no, I can definitively say we could not have done it in Python at all. And we are very, very lucky to benefit from a lot of the open source contributors that have, you know, for free, devoted their time and efforts to bringing support for languages like Rust. I'm imagining that another, you know, timing tailwind is the availability of coding agents now.

[20:38](https://realalpha.app/v/MAABmDB51LY/?t=1238) You know, you may have needed to build a lot of, you know, supporting infrastructure before because it just wasn't available for what you're trying to do. But now you can just ask an agent to build that for you. For sure. I mean, a thousand percent. You mentioned that Wayfair was an early adopter of transformers. Are your models transformer based? We have some that are transformer based. We have, I would say, all of our models certainly use self-attention, the self-attention mechanism.

[21:06](https://realalpha.app/v/MAABmDB51LY/?t=1266) I can say that without a doubt. All of them do. We have many mixture of experts models now, but they all rely on self-attention. I wouldn't say that they use the transformer architecture across the board because, I mean, now we have so many proprietary models that we have built that kind of it's kind of like an orchestra of many many many many models it's like mixture of experts but for ranking is the idea