Software Engineer, Platform

Software is eating the world, but AI is eating software. We live in unprecedented times – AI has the potential to exponentially augment human intelligence. Every person will have a personal tutor, coach, assistant, personal shopper, travel guide, and therapist throughout life. As the world adjusts to this new reality, leading platform companies are scrambling to build LLMs at billion scale, while large enterprises figure out how to add it to their products. To make them safe, aligned and actually useful, these models need human eval and reinforcement learning through human feedback (RLHF) during pre-training, fine-tuning, and production evaluations. This is the main innovation that’s enabled ChatGPT to get such a large headstart among competition.

At Scale, our products include the Generative AI Data Engine, SGP, Donovan, and others that power the most advanced LLMs and generative models in the world through world-class RLHF, human data generation, model evaluation, safety, and alignment. The data we are producing is some of the most important work for how humanity will interact with AI.

At the foundation of these products is the SGP Platform Engineering team. In this role, you will support the design and development of shared platforms used across Scale. This includes designing our foundational data platforms and lifecycle, architecting Scale’s core cloud infrastructure and orchestration stack, and redefining how engineers develop, build, test, and deploy software at Scale. You’ll also get widespread exposure to the forefront of the AI race as Scale sees it in enterprises, startups, governments, and large tech companies.

You will:

• Drive the design and implementation of our foundational platforms and systems that will become the baseline for Scale’s capabilities.
• Collaborate with cross-functional teams to define and understand requirements, design, and deliver new features within a highly iterative and innovative software development lifecycle.
• Work within the team to lead the market in your team’s AI domain through experimentation with the latest techniques and technologies in this area.
• Proactively identify opportunities for and drive improvements to current programming practices, including process enhancements and tool upgrades.
• Present technical information to teams and stakeholders, providing guidance and insights on development processes and technologies.

Ideally you’d have:

• 5+ years of full-time engineering experience, post-graduation, with specialties in back-end systems.
• Excitement to work with AI technologies and build agentic experiences.
• Extensive experience in software development and a deep understanding of highly scalable and reliable distributed systems hosted on public cloud platforms.
• Experience building document retrieval and understanding systems, such as RAG, to accurately ingest large-scale client knowledge into client agents and applications.
• A strong track record of independently owning and delivering successful engineering projects, including the ability to collaborate with and inspire others to engage with the problem space.
• Excellent communication and collaboration skills, and the ability to translate complex technical concepts to non-technical stakeholders.
• Experience working fluently with standard containerization and deployment technologies like Kubernetes, Terraform, Docker, etc.
• Experience with long-running agent/code orchestration platforms, such as Temporal.
• Mastery of storing state with structured databases, including Postgres.
• Strong knowledge of software engineering best practices and CI/CD tooling (CircleCI, GHA).

Nice to haves:

• Experience across multiple cloud platforms (GCP, AWS, Azure, Oracle, on-premise).
• Experience with multiple ways to persist contextual data, e.g., knowledge graphs and hierarchical indexes.
• Experience with other types of large data solutions, e.g., Databricks and Snowflake.
• Experience with authentication/authorization systems (Zanzibar, Authz, etc.).
• Experience scaling technical products at hyper-growth startups.