---
title: 'Machine Learning Engineer, Voice at Speak'
canonical: 'https://feeny.ai/job/machine-learning-engineer-voice-speak-san-francisco-58kkpcqzq1kk'
type: 'job'
last_seen: '2026-09-13'
---

# Machine Learning Engineer, Voice at Speak

- **Company:** Speak
- **Location:** San Francisco, CA
- **Compensation:** $200k–$300k
- **Employment:** full-time
- **Work type:** hybrid
- **Posted:** 2024-03-05
- **Last confirmed live:** 2026-09-13
- **Apply:** https://jobs.ashbyhq.com/speak/e78edff4-5135-4c68-932d-e449ee460ea3/application

**Skills:** Python, PyTorch, Deep Learning, GPU Programming, Machine Learning Pipelines, Speech Recognition, ASR, Speech Audio Processing

> Develop and deploy end-to-end speech recognition models for language learning apps. Responsibilities include training large models on GPUs, improving pronunciation feedback systems, expanding multilingual support, and building data infrastructure for training and evaluation.

## Job description

## About us

Our mission is to reinvent the way people learn, starting with language. Learning a language can change a life by opening doors to new cultures, careers, and communities. Two billion people around the world are actively trying to learn a language, but the best way to learn (one-on-one tutoring) is hard to access at scale and hasn’t been meaningfully improved in decades. Speak is building a human-level, AI-powered tutor in your pocket: a conversation-first experience that lets learners actually speak, get instant feedback, and progress through carefully designed lessons. The result is a complete path from beginner to confident speaker across multiple languages. Speak first launched in South Korea in 2019, where Speak has now become the number one language learning app, and we now serve learners across many markets and 15+ languages. Speak is one of the world’s leading AI companies, with over $150m raised in venture investment from OpenAI, Accel, Founders Fund, Khosla Ventures, and more, with a distributed team across San Francisco, Seoul, Tokyo, Taipei, and Ljubljana.

## About this role

We are looking for an experienced Machine Learning Engineer to join our team and help develop cutting-edge speech recognition models that help teach language fluency. In this role you will take ownership of the end-to-end modeling pipeline, from training and experimentation to deployment and monitoring. You will also work closely with Product teams to design innovative learning experiences and measure the efficacy of production models as they affect our end users. We are a small, dynamic team where you will contribute as a developer and thought partner on team projects like ASR, assessment, pronunciation, content personalization, and much more. This is an incredibly exciting time to join an ML team designing a personalized learning experience that will revolutionize language learning for millions of learners worldwide — come join us!

## What you'll be doing

- Training and deploying ASR models end-to-end, including monitoring, performance tracking, and retraining
- Improving the pronunciation model that provides precise feedback, and make it more central to our learning app
- Creating metrics to measure ASR performance across tasks and languages
- Expanding our ASR systems to new languages and markets
- Building and maintaining data infrastructure such as training/evaluation datasets and labeling pipelines

## What we're looking for

- Extensive experience training large models on GPUs and deploying custom deep learning models
- Proficiency in Python and common Deep Learning frameworks like PyTorch
- Demonstrated experience owning ML pipelines end to end, from POC to production
- Strong communication skills and the ability to explain complex ML concepts to non-technical stakeholders
- Sharp product sense and an ability to think broadly and cross-functionally about model quality in the context of user experience
- Bonus
- Experience with speech or audio

Office

- San Francisco, CA

## Why work at Speak

- Join a fantastic, tight-knit team at the right time: we're growing very quickly, we've most recently raised our Series C from some of the top investors in the valley, and we've achieved product-market fit in our initial markets. You'd join at a magical time when a single person could significantly change the course of the company.
- Do your life's work with people you’ll love working with: we care strongly about our craft and want every person at Speak to feel like they're growing every day. We believe in the idea that working with people you both enjoy and have respect for makes everything better. We hire thoughtfully and only work with people we admire deeply.
- Global in nature: We're live in over 40 countries and launching in a number of new markets soon. We have dedicated offices in San Francisco, Ljubljana, Seoul, Tokyo, and Taipei, and you’ll have the opportunity to talk to users in each of these regions on a regular basis as well as travel.
- Impact people's lives in a major way: Learning a language is one of the single most life-changing skills one can learn, and right now 99% of people never achieve their goal because the process is broken. We’re helping millions of people achieve their goals and improve their lives.

Speak does not discriminate based upon race, religion, color, national origin, gender (including pregnancy, childbirth, or related medical conditions), sexual orientation, gender identity, gender expression, age, status as a protected veteran, status as an individual with a disability, or other applicable legally protected characteristics.

## About Speak

## Company Overview
- **One-liner**: Speak is an AI-powered language learning app that focuses on teaching conversational fluency through real-time speaking practice and instant feedback, without needing a human tutor.
- **Entity Type**: Private (Series C; $1B valuation as of December 2024)
- **Headquarters**: San Francisco, California, USA (with offices in Seoul, Tokyo, Taipei, and Ljubljana)
- **Founded**: 2016
- **Founders**: Connor Zwick (CEO) and Andrew Hsu (CTO)

## Core Business
- **Industry**: Language learning / EdTech
- **Target Customers**: B2C (individual language learners) and B2B (Speak for Business – over 200 enterprise customers)
- **Mission/Purpose**: “To reinvent the way people learn, starting with language.” The goal is to make conversational practice radically accessible so that hundreds of millions of people can gain fluency.

## Products & Services
- **Speak App**: A subscription-based mobile and web application (iOS, Android, Web) that uses a proprietary “Speak Method” – learn useful phrases, practice them out loud with AI-driven drills, then apply them in simulated real-world conversations. Supports English learning from eight source languages; plans to add Spanish and French.
- **Speak for Business**: A B2B tier for companies that want to provide English conversation training to their employees, currently serving 200+ corporate clients.

## Market Standing
- **Valuation**: $1 billion (December 2024, Series C led by Accel)
- **Key Metric**: Total funding of $171.3 million across six rounds; annual revenue estimated at $7.0 million (LinkedIn data)
- **Notable Investors**: Accel, OpenAI Startup Fund, Founders Fund, Khosla Ventures, Y Combinator, Lachy Groom, Buckley Ventures; individual angels include Sam Altman and Peter Thiel.
- **Growth Signals**: 
  - 10M+ app downloads, #1 English learning app in South Korea
  - Available in over 30 countries
  - Headcount grew 30.9% YoY to 166 employees (LinkedIn, May 2026)
  - B2B segment has 200+ customers
  - Average user engagement of 10–20 minutes per day

## Competitive Advantages
- **Conversation-first approach**: Unlike grammar/vocab apps, Speak teaches by making users speak out loud, using proprietary speech recognition and generative AI to provide immediate feedback.
- **AI tutor that scales**: Replicates natural conversation without needing a live human, making high-quality practice affordable ($20/month or $99/year).
- **Deep OpenAI partnership**: Early access to the latest speech AI from OpenAI (also an investor), giving a technological edge in voice interaction.
- **Strong product-market fit in Asia**: Proven adoption in Korea, Japan, Taiwan, and expanding to other markets.

## Strategic Focus
- Expand target languages (Spanish and French are next)
- Scale Speak for Business globally
- Develop a standardized, accurate English fluency assessment (an AI-powered test)
- Continue to invest in AI research and engineering to improve tutoring effectiveness

## Why Work Here
- **Culture**: Mission-driven team passionate about language learning and AI; values efficacy over gamification.
- **Team & talent**: 166 people from companies like Meta, Google, DoorDash, Block, Dropbox, and Duolingo. Engineering office in Ljubljana, product and design hubs in San Francisco, operations in Seoul, Tokyo, Taipei.
- **Flexibility**: Hybrid/office-based across locations; employees encouraged to travel between offices; annual company offsite.
- **Perks**: Competitive salary and equity, comprehensive medical/dental/vision, unlimited PTO, parental leave, 401(k), monthly wellness and language learning stipends, pet-friendly office (official company dog).
- **Growth**: Rapidly scaling headcount (+30% YoY) with active roles in engineering, product, marketing, sales, and design.

## Sources
1. [speak.com/careers](https://www.speak.com/careers)
2. [speak.com](https://speak.com/)
3. [linkedin.com/company/usespeak](https://www.linkedin.com/company/usespeak)
4. [TechCrunch: OpenAI-backed Speak raises $78M at $1B valuation](https://techcrunch.com/2024/12/10/openai-backed-speak-raises-78m-at-1b-valuation-to-help-users-learn-languages-by-talking-out-loud/)
5. [Y Combinator: Speak](https://www.ycombinator.com/companies/speak)

## Other roles at Speak

- [Product Lead, Enterprise](https://feeny.ai/job/product-lead-enterprise-speak-san-francisco-pv45a3bsw3js) — San Francisco, CA
- [Content Localization Specialist (Chinese - Traditional / Japanese)](https://feeny.ai/job/content-localization-specialist-chinese-traditional-japanese-speak-south-korea-9c7ckjwm39cd) — South Korea
- [Assessment Design Lead](https://feeny.ai/job/assessment-design-lead-speak-united-states-w55yxvzn27x2) — United States
- [Full-stack Engineer](https://feeny.ai/job/full-stack-engineer-speak-san-francisco-z8ecxdk1xw63) — San Francisco, CA
- [B2B Sales Lead - Korea](https://feeny.ai/job/b2b-sales-lead-korea-speak-seoul-k7qx1kfghnya) — Seoul, South Korea
- [People Ops Specialist (Contract) - Korea](https://feeny.ai/job/people-ops-specialist-contract-korea-speak-seoul-wrbbd36m9z8c) — Seoul, South Korea
- [Content Localization Specialist (Japanese / Korean)](https://feeny.ai/job/content-localization-specialist-japanese-korean-speak-south-korea-kenmfgqkdwrf) — South Korea
- [Content Localization Specialist (Korean / Chinese - Simplified)](https://feeny.ai/job/content-localization-specialist-korean-chinese-simplified-speak-south-korea-9qvr499kdcq7) — South Korea
- [Lead Business Analyst](https://feeny.ai/job/lead-business-analyst-speak-san-francisco-dzkeyxkp1096) — San Francisco, CA
- [Senior Accountant](https://feeny.ai/job/senior-accountant-speak-san-francisco-h4wye00jhcc6) — San Francisco, CA