Software Engineering Intern (Summer 2027)

Public source summary from intern-list.com / Jobright.

Niantic Spatial is building physical AI and spatial intelligence technologies for understanding, representing, navigating, and engaging with the real environment. Software engineering interns will own production software projects in specialized ML/AI infrastructure, product engineering, or backend systems tracks, working on 3D reconstruction, foundation models, spatial data, APIs, and positioning infrastructure.

Responsibilities

Model Training with PyTorch — Implement, train, and evaluate model architectures in PyTorch, iterating on data loaders, loss functions, and training loops to improve model quality and convergence on real production datasets Training Infrastructure — Design and scale distributed training pipelines for our Large Geospatial Model, handling petabyte-scale spatial data across multi-GPU and multi-node environments Data Pipelines — Build high-throughput ingestion and preprocessing pipelines that transform raw imagery and 3D point clouds into training-ready datasets Observability — Instrument training runs with metrics, dashboards, and alerting so engineering teams can debug and iterate faster Product Features — Build and ship features for managing organizations, projects, spatial data, and content API Design — Design the interfaces that connect our products, web and mobile applications, and developer tools, making them consistent, well-documented, and easy to use Workflows & Integrations — Connect the steps behind core product workflows, from uploading and processing data to making results available to users and applications End-to-End Delivery — Collaborate with engineers across frontend, backend, and infrastructure to launch features and learn from real customer use Service Development — Design and ship microservices in Go or C++ that sit in the critical path of our positioning and reconstruction APIs Data Ingestion — Build and optimize pipelines that ingest, validate, and route large volumes of visual and sensor data from diverse hardware sources Performance & Reliability — Profile service bottlenecks, improve latency, and improve system observability through structured logging, metrics, and distributed tracing API Design — Contribute to internal and external API design, writing clean, well-tested, production-grade interfaces that other teams and customers depend on

Qualifications: Currently pursuing a BS or MS in Computer Science, Robotics, Electrical Engineering, Systems Engineering, Computer Vision, or a related field TypeScript, Python, C++ or Go experience for Infrastructure and Backend tracks Genuine curiosity about how AI interacts with the physical world - 3D reconstruction, spatial reasoning, or real-world positioning systems Ability to work independently, debug ambiguous problems, and communicate clearly with a small team Students obsessed with the intersection of ML and systems - PyTorch, distributed computing (Ray, Spark), CUDA, and high-performance architecture Students who enjoy building products from the backend up and interested in how APIs and data power user experiences.

Motivated by shipping software that real customers and developers use Pragmatic engineers who care about correctness, performance, and clean systems design and want to see their code serving real traffic within weeks Experience with cloud infrastructure - Kubernetes, AWS or GCP, Docker, Terraform Familiarity with CUDA or GPU parallelization Open-source contributions to libraries like PyTorch, OpenCV, COLMAP, or similar Prior work in robotics, autonomous systems, XR, or spatial computing (coursework, research, or projects acceptable) Exposure to Gaussian Splatting, NeRF, or 3D reconstruction techniques Familiarity with REST APIs Experience with SQL or other relational databases Experience using AI-assisted development tools (Claude, ChatGPT, etc.) to write, debug, and improve code

Benefits: Housing assistance is available for candidates outside the Bay Area - details provided during the offer process. We provide lunch every day. Snacks are always stocked.