TL;DR
Founding Product Engineer (AI): Building an AI interviewer that turns qualitative conversations into quantitative insights with an accent on turning free-form conversation into structured data using embeddings and fine-tuned LLMs. Focus on solving complex challenges in natural language understanding, LLM prompting, and multi-modality speech-to-speech pipelines for scalable and insightful user interactions.
Location: Remote, with office available in San Francisco.
Salary: $150,000–$200,000 USD in cash
Company
hirify.global has built an AI interviewer that can ask follow-up questions, like a conversation, but at the scale and ease of use of a survey.
What you will do
- Develop systems to turn qualitative conversation data into quantitative insights using embeddings and fine-tuned large language models.
- Implement novel fine-tuning and prompting techniques to enable LLMs to ask the right questions in conversations.
- Improve and optimize multi-modality speech-to-speech pipelines for robust audio input and output with LLMs.
- Work on products end-to-end, taking full ownership of the user experience.
- Collaborate with other engineers to build scalable and maintainable systems.
Requirements
- Experience working on products end-to-end without detailed direction.
- Proficiency with Next.js, TypeScript, and large language models.
- Strong ability to build systems collaboratively with other engineers in mind.
- Excellent written and verbal communication skills.
- Highly technical with an interest in building products end-to-end and owning the user experience.
- Excitement about large language models.
Culture & Benefits
- High equity grant, typical for an early-stage startup.
- Opportunity to work on cutting-edge AI challenges in a collaborative environment.
- Access to an office @ 425 2nd St in San Francisco for in-person work, if desired.
Будьте осторожны: если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого - это мошенники. Обязательно жмите "Пожаловаться" или пишите в поддержку. Подробнее в гайде →