
Turbopuffer is looking for a Research Engineer who wants to build the next generation models for search.
What You'll Do
- Work with our inference team to integrate and evaluate the best embedding and reranking models.
- Design, train, and evaluate state-of-the-art models that integrate closely with the database.
- Communicate internally and externally about model choices, evaluations, and trade-offs.
- Automate yourself out of a job.
Requirements
Research
- Extensive experience with embedding/reranker-model pre-training and/or fine-tuning.
- Proven experience conducting large-scale, distributed training runs.
- Ability to work with and reason about large datasets.
- Strong communication skills to present research findings to audiences outside the field and guide customers in model choices.
- Passion for shipping research to production.
Domain Expertise
- Opinionated take on model benchmarks and evaluation.
- Up-to-date knowledge of state-of-the-art retrieval and search papers.
- Ability to quickly iterate on new research ideas to verify or debunk hypotheses.
Soft Skills & Mindset
- Writer: Ability to write well and explain complex technical ideas simply.
- Integrity: Build trust and comfortably admit what you don't know.
Nice to Have
- Experience with model inference and managing GPU infrastructure.
- Experience working with systems languages (Rust, C++, Java, Go, etc.).
- Prior experience in vector and full-text search systems.
Values
- Overstep > understep
- Show > tell
- Complexity is earned
- Showing up as part of our customerβs team
- Making stuff easy to react to
Timezone overlap
UTC-8β-4
Open to
NA
Sign in to track applications and earn points.