---
title: 'Machine Learning Scientist at Relace'
canonical: 'https://feeny.ai/job/machine-learning-scientist-relace-san-francisco-x6cd3a3vkg65'
type: 'job'
last_seen: '2026-09-11'
---

# Machine Learning Scientist at Relace

- **Company:** Relace
- **Location:** San Francisco, CA
- **Employment:** full-time
- **Work type:** onsite
- **Posted:** 2025-10-29
- **Last confirmed live:** 2026-09-11
- **Apply:** https://jobs.ashbyhq.com/relace/bc8f030e-157f-44b1-95ac-65fb1d807cc1

## Job description

## About Us

Relace is building the models and infrastructure that code agents reach for. We power the fastest model on OpenRouter (10,000 tok/s) and deliver optimized small language models designed for retrieval, application, and core code generation functions. Our technology supports some of the world’s fastest-moving companies — including Lovable, Figma, and Vercel — as they deploy and scale code generation to hundreds of millions of users. We recently raised our Series A from a16z, and we’re growing quickly. Our team is made up of mathematicians, physicists, and computer scientists who are deeply passionate about their craft. If you thrive on ambitious technical problems, care about elegant systems design, and want to build the foundation of how code gets written at scale, this is the place for you.

## The Role

We’re looking for a Machine Learning Scientist to push the limits of small, high-performance language models. This is a deeply technical role focused on advancing the capabilities of our models for retrieval, application, and code generation. The ideal candidate has a strong background in ML research and engineering, is comfortable working with both theory and production systems, and thrives in an environment where ideas turn into deployed infrastructure fast. This person should be excited to work on training methodology, optimization, evaluation, and model architecture at scale — and collaborate directly with infrastructure and product teams to get breakthroughs into production quickly. This role is best suited for someone who loves both mathematical elegance and real-world impact.

## Requirements

- Strong background in machine learning, deep learning, or related fields.
- 2+ years of experience working on ML research or production systems.
- Fluency in Python and frameworks like PyTorch or JAX.
- Experience with training and optimizing large or efficient models.
- Strong understanding of applied optimization, distributed training, or model evaluation.
- Familiarity with code models, retrieval systems, or language modeling a plus.
- Advanced degree (MS or PhD) in a quantitative field, or equivalent industry experience.
- Willingness to work in-person from our SF office in FiDi.

## About Relace

## Company Overview
- **One-liner**: Relace builds specialized AI models and infrastructure for coding agents, enabling fast, cheap, and reliable code retrieval and merging.
- **Entity Type**: Private (startup; Y Combinator Winter 2023)
- **Headquarters**: San Francisco, California, USA
- **Founded**: 2022
- **Founders**: Preston Zhou and Eitan Borgnia

## Core Business
- **Primary industry**: AI infrastructure for software development (coding agents, code generation)
- **Target customers**: B2B – AI codegen startups, engineering teams, and enterprises integrating autonomous coding agents into their workflows.
- **Mission**: Build the models and infrastructure to make coding agents a seamless part of all software systems – turning static SaaS pages into malleable user-driven interfaces.

## Products & Services
- **Instant Apply**: A universal code merging model that applies file edits at >10,000 tokens per second (end-to-end latency ~900ms). Reduces token usage by ~40% compared to full-file rewriting with frontier LLMs. [relace.ai](https://relace.ai/)
- **Embedding + Code Reranker**: Semantic search that retrieves relevant context from million-line codebases in ~1–2 seconds, cutting input token usage by over 50% and improving generation quality. [ycombinator.com](https://www.ycombinator.com/companies/relace)
- **Source Control for Agents**: Lightweight push/pull operations, fast branching for subagents, automatic indexing for two-stage retrieval, and rate limits designed for high throughput. [relace.ai](https://relace.ai/)
- **Self-Hosted / VPC Deployments**: On-premise or VPC-isolated deployments for teams with strict compliance or latency requirements. [relace.ai](https://relace.ai/)

## Market Standing
- **Valuation**: Not publicly disclosed (latest round: $23M raised as of Oct 2025 per YC news) [ycombinator.com](https://www.ycombinator.com/companies/relace)
- **Key Metric**: Total funding – $23M (announced October 2025)
- **Notable Investors/Partners**: Y Combinator (Winter 2023 batch), plus production partnerships with Lovable, Magic Patterns, Codebuff, Create, Tempo Labs, and 20+ other AI codegen startups. [ycombinator.com](https://www.ycombinator.com/companies/relace)
- **Growth Signals**: Models running millions of times per week in production; SOC 2 compliant; team of 7 (as of YC profile); hiring across multiple roles (ML, engineering, GTM). [builtin.com](https://builtin.com/company/relace)

## Competitive Advantages
- **Purpose-built SLMs**: Unlike general-purpose LLMs, Relace trains specialized small language models (SLMs) for retrieval and merging – achieving order-of-magnitude improvements in latency and cost without sacrificing accuracy. [relace.ai](https://relace.ai/)
- **Co-optimized infrastructure**: Models and infrastructure are designed together for coding agent workflows, enabling 10k+ tok/s merging and sub-second codebase search. [ycombinator.com](https://www.ycombinator.com/companies/relace)
- **Strong adoption**: Already integrated by leading AI codegen platforms, proving real-world performance and reliability.

## Strategic Focus
- **Expand model capabilities**: Continue training specialized SLMs for additional coding agent tasks (e.g., debugging, testing).
- **Deepen enterprise readiness**: Self-hosted/VPC deployments and SOC 2 compliance signal a push into larger, compliance-sensitive customers.
- **Build the “rails for software on-demand”**: Enable coding agents to be embedded into any software system, making interfaces malleable and user-driven.

## Why Work Here
- **Culture**: In-person team in San Francisco; described as “ex-academics, founders, and hackers” who are “stubbornly optimistic” about solving hard technical problems. [relace.ai/about-us](https://relace.ai/about-us)
- **Work policy**: In-office (San Francisco) with a hybrid workspace option for some roles. [builtin.com](https://builtin.com/company/relace)
- **Engineering focus**: High agency, early-stage environment where engineers build core AI models and infrastructure from scratch. Roles include Machine Learning Scientist, Machine Learning Engineer, Infrastructure Engineer, Product Engineer (Frontend), and GTM Engineer. [jobs.ashbyhq.com/relace](https://jobs.ashbyhq.com/relace)
- **Impact**: Work on models used millions of times per week by leading AI codegen startups – direct influence on the future of software development.

## Sources
1. [relace.ai](https://relace.ai/)
2. [relace.ai/about-us](https://relace.ai/about-us)
3. [ycombinator.com/companies/relace](https://www.ycombinator.com/companies/relace)
4. [builtin.com/company/relace](https://builtin.com/company/relace)
5. [jobs.ashbyhq.com/relace](https://jobs.ashbyhq.com/relace)

## Other roles at Relace

- [Product Engineer (Frontend)](https://feeny.ai/job/product-engineer-frontend-relace-san-francisco-s47yehc0311k) — San Francisco, CA
- [Machine Learning Engineer](https://feeny.ai/job/machine-learning-engineer-relace-san-francisco-krnc269vmpmj) — San Francisco, CA
- [Developer Relations Advocate](https://feeny.ai/job/developer-relations-advocate-relace-san-francisco-vxvkyvpq0gwt) — San Francisco, CA
- [Growth & Marketing Engineer](https://feeny.ai/job/growth-marketing-engineer-relace-san-francisco-pjhd6kdw85td) — San Francisco, CA
- [Infrastructure Engineer](https://feeny.ai/job/infrastructure-engineer-relace-san-francisco-zf8xjm22pchz) — San Francisco, CA
- [Machine Learning Scientist](https://feeny.ai/job/machine-learning-scientist-flagship-pioneering-inc-somerville-ma-98f7vrcztgjm) — Somerville MA, United States
- [Machine Learning Scientist](https://feeny.ai/job/machine-learning-scientist-eli-health-montreal-b0xn6yyc72ff) — Montréal, Canada
- [Machine Learning Scientist](https://feeny.ai/job/machine-learning-scientist-tacit-san-francisco-pt1w314e43ge) — San Francisco, CA
- [Machine Learning Scientist](https://feeny.ai/job/machine-learning-scientist-latent-labs-london-9m8twmtyrvpj) — London, United Kingdom
- [Machine Learning Scientist](https://feeny.ai/job/machine-learning-scientist-spotter-culver-city-california-wgbw208cbr43) — Culver City California, United States