Software Engineer I (Coding)

Handshake•San Francisco, CA

23h•Onsite

About The Position

As a Software Engineer on Handshake's Coding Pod, you will be responsible for building the data infrastructure and pipelines that power frontier AI coding models. This team operates at the intersection of applied machine learning, data engineering, and product development, focusing on creating large-scale, high-quality benchmark datasets for evaluating model performance on real-world coding tasks. You will own the entire data pipeline lifecycle, from task generation and dataset construction to quality assurance and delivery, playing a key role in defining how next-generation coding models are trained and evaluated. The role involves working with human-in-the-loop evaluation systems, agentic coding tools, and automated assessment frameworks to ensure dataset quality and model performance. This position is ideal for engineers who enjoy working across backend systems, data pipelines, and developer tooling, and are motivated by the prospect of directly impacting the future of AI-powered software development.

Requirements

2+ years of professional software engineering experience, with exposure to backend systems or data engineering.
Strong programming skills (e.g., Python, TypeScript, or similar).
Experience building or maintaining data pipelines, ETL systems, or distributed systems.
Familiarity with containerization (e.g., Docker) and cloud infrastructure (AWS, GCP, or similar).
Understanding of databases and data modeling (SQL or NoSQL).
Comfort working across systems boundaries, from infrastructure to developer-facing tools.
Strong problem-solving skills and attention to detail, especially in data quality and correctness.
Effective communication skills and ability to collaborate cross-functionally.

Nice To Haves

Experience working with machine learning datasets, evaluation frameworks, or benchmarking systems.
Familiarity with coding agents, developer tools, or AI-assisted programming systems.
Experience integrating with APIs like GitHub or building developer platform tooling.
Exposure to workflow orchestration tools (e.g., Airflow, Temporal) or distributed job systems.
Experience designing automated testing or grading systems for code.
Background in high-growth or infrastructure-heavy engineering environments.
Interest in the intersection of AI, developer productivity, and real-world software engineering workflows.

Responsibilities

Design and build scalable data pipelines for generating, transforming, and validating large-scale coding datasets.
Develop systems for task generation, dataset curation, and quality assurance, including automated and human-in-the-loop evaluation workflows.
Integrate with developer ecosystems (e.g., GitHub) and build tooling that supports real-world coding environments.
Work with containerized environments (e.g., Docker) to safely execute and evaluate code at scale.
Build backend systems and APIs that power dataset delivery and model evaluation pipelines.
Collaborate closely with ML researchers, product managers, and other engineers to define evaluation methodologies and improve dataset quality.
Implement automated grading, benchmarking, and assessment systems for coding tasks.
Debug and optimize pipeline performance, reliability, and scalability across distributed systems.
Contribute to architectural decisions around data infrastructure, evaluation systems, and pipeline orchestration.