AI Product Engineer, Speak
2-3 days a week in office
Language learning app
Open for applications
Language learning app
201-500 employees
Open for applications
2-3 days a week in office
201-500 employees
Company mission
To reinvent the way people learn, starting with language.
Role
Who you are
- 3-5+ years of experience in full stack/backend, product-focused software engineering
- Proficiency in React/Node/Typescript and Python
- Strong product intuition — the ability to think broadly and cross-functionally about innovative LLM-powered capabilities and product experiences
- Ability to work independently and build at a high velocity
- Interest in building with LLMs
- You may have participated in LLM hackathons, developed several side projects, experimented with different types of models, and use ChatGPT daily
Desirable
- You have real-world experience developing and deploying LLM apps and a strong understanding, gained through experience, of what works and what doesn't
- A keen intuition for improving performance and output quality of LLM systems
- Experience with LLM Ops and tools (e.g., vector databases, RAG, prompt ops)
What the job involves
- As an AI Product Engineer at Speak, you'll play a pivotal role in developing the future of language learning and creating the most effective path to language fluency
- Your primary responsibility will be developing innovative product experiences powered by Language Learning Models (LLMs) and deploying them to millions of users
- Your tasks will span the full stack and beyond, from evaluating models and testing prompts to developing and iterating on various product features such as our conversational onboarding and lesson experiences, grammatical assessment and feedback, and personalized lessons
- We're constantly pushing the boundaries of what LLMs can do to provide an exceptional and unparalleled language learning experience to users in over 30 countries worldwide
- Developing and deploying LLM-powered language learning products across the full stack, as well as enhancing the quality and performance of existing AI-powered features within Speak
- Collaborating cross-functionally with other Engineering teams, Applied ML, Product, Design, and Content
- Refining our process for building LLM apps, including best practices for prompting, experimentation/evaluation, LLM Ops, measuring quality and performance, etc
- Scaling existing product features to many more users and languages
Our take
Language is one of the most powerful skills a person can learn. It helps us connect with people, understand different cultures, travel with confidence, and unlock new opportunities in life and work. But learning a language is often easier said than done. Traditional lessons are daunting and slow, not to mention expensive, and many people never get enough practice actually speaking out loud.
Speak is changing that by putting an AI language tutor right in your pocket. Instead of focusing on memorising endless vocabulary lists, Speak is built around real conversations. Learnings speak from the very beginning, receive instant feedback, and work through lessons designed to help them improve naturally over time. It brings the experience of practicing with a personal tutor to millions of people, making language learning more accessible.
That approach is clearly working too. Since launching in South Korea in 2019, Speak has now grown into the country's number one language learning app and now supports learners across more than 15 languages worldwide. Backed by leading investors, including OpenAI, Speak has become on of the world's leading AI language learning companies. And as AI becomes better at understanding and responding like a real conversation partner, Speak is in a good position to help even more people not just learn a language but become confident in it.
Kirsty
Company Specialist at Welcome to the Jungle
Company
Funding (last 2 of 6 rounds)
Dec 2024
$78m
SERIES C
Jun 2024
$20m
SERIES B
Total funding: $165.9m
Company benefits
- Competitive salary and equity
- Comprehensive medical, dental, and vision insurance
- Unlimited PTO
- Parental leave
- 401(k)
- Monthly wellness stipend
- Monthly language learning stipend
Company HQ
Potrero Hill, San Francisco, CA
Leadership
Connor Zwick
(Co-Founder & CEO)Prior to co-founding Speak in 2016, Connor founded Flashcards+ where they served as the CEO until it was acquired by Chegg. They also served as the CEO at Coco Controller for nearly 3 years.
Salary benchmarks
We don't have enough data yet to provide salary benchmarks for this role.
Submit your salary to help other candidates with crowdsourced salary estimates.
Share this job
View 13 more jobs at Speak