Frontier Data Lead - Code

Turing•San Francisco, CA

2d•Remote

About The Position

Turing is the world’s leading research accelerator for frontier AI labs and a trusted partner for global enterprises looking to deploy advanced AI systems. Turing accelerates frontier research with high-quality data, specialized talent, and training pipelines that advance thinking, reasoning, coding, multimodality, and STEM. For enterprises, Turing builds proprietary intelligence systems that integrate AI into mission-critical workflows, unlock transformative outcomes, and drive lasting competitive advantage. The Role We create RL environments to evaluate and improve our customers' models on complex, long-range, multi-step workflows across high-GDP-value domains such as Finance, Sales, Retail, Developer Tools, Collaboration, Customer Experience. The environments vary depending on the model capability being evaluated / improved, a few examples of environment types are listed here: Environments for Software Engineering / coding agents UI-Environments for Computer-Use/Browser-Use agents MCP-based Environments for general function-calling agents across various enterprise and consumer applications The Frontier Data Lead – Code will own end-to-end the creation of datasets, RL environments, and evals for frontier AI labs in the domain of coding agents and software engineering. This is a hands-on technical leadership role where you influence revenue directly – you will be mapped to one or more AI labs and interface directly with researchers / engineers at those labs to understand their needs and build data offerings to address those needs. To achieve this, you will build and manage teams of software engineers, researchers, QAs, and contractors/data-annotators from Turing’s talent pool of 4M+ developers. You’ll be responsible for delivering projects at frontier quality and scale—owning data quality, throughput, and timely delivery. You’ll define and manage data pipelines, validation workflows, and review processes to ensure datasets meet the highest standards for realism, correctness, and diversity. You’ll also develop automations, synthetic data generation systems, and internal tools to scale production efficiently. In short, you’ll run your project like a startup within Turing, owning both the technical architecture and the operational execution required to produce best-in-class datasets/environments/evals to make the world’s best coding agents and models even better at real-world coding tasks across the software development lifecycle.

Requirements

Post-training experience on SWE tasks or experience building coding agents: We expect that you have a deep understanding of data ingredients and design principles that lead to measurable coding model improvements, either from fine-tuning models to improve SWE capabilities or building your own coding agents to improve upon SWE capabilities of the base model.
Engineering Management experience: have led teams of engineers in the past, including interviewing/hiring them and setting up QA processes.
Hands-on technical capability: Fluency in Python and proficiency in one or more major languages (C++, Java, Go, Rust, or JS).
Operational leadership: Proven ability to manage complex data pipelines, multi-stakeholder delivery, and concurrent high-stakes projects.
Cross-functional communicator: ability to communicate clearly with researchers at frontier AI labs, subject matter experts for various domains, and diverse teams.
Background in Computer Science, Machine Learning, or related technical field preferred.

Responsibilities

Lead the creation of datasets, rl environments, and evals focused on Coding Agents / Software Engineering for one or more AI lab customers.
Ensure that everything you ship to clients meets frontier standards for realism, correctness, diversity, and difficulty.
Set up quality rubrics, automated validation scripts, and human review processes for every stage of data generation.
Build and lead cross-functional teams of software engineers, researchers, QAs, and data creators drawn from Turing’s 4M+ developer network.
Interview, onboard, train, and mentor team members to ensure consistent output quality and technical excellence.
Act as the primary technical point of contact for your customer projects, interfacing directly with researchers and engineers at frontier AI labs to understand their coding agent roadmap and model data needs, to gather feedback, and to co-define success criteria for your projects.
Provide regular progress updates, surface insights from model evaluations, and incorporate client feedback to improve future iterations.
Fine-tune models in-house on Turing-generated datasets or Turing-rl-environment generated trajectories to determine model improvement as a proof of data quality
Proactively build benchmarks and run evals on frontier models and coding agents to identify strengths and weaknesses on SWE tasks, and leverage these insights to inform product roadmap
Equip customer-facing teams with the Evaluation reports, sample datasets, and trainings to enable them to communicate your data offerings to customers most effectively
Publish research papers and technical posts on Turing’s data products, innovations in our synthetic data generation / automation pipelines, evaluations of frontier agents and models, and Turing’s model fine-tuning results on our datasets.
Oversee development of internal tools that accelerate data generation and verification (e.g., automated data scraping pipelines, unit test generators, repo sandboxing).
Design dashboards and APIs for customers to run model evals, view performance reports, and integrate Turing data directly into their post-training pipelines.