Research Engineer

Breakout Ventures; San Francisco (CA); On-site; USD 180,000 - 250,000; Full time

Benefits offered by this job

Health insurance

Relocation assistance

Workspace stipend

Team events

Job summary

Constellation is seeking a Research Engineer in San Francisco to bridge data and models, accelerating long-horizon multimodal training. You will build data pipelines, libraries, and evaluation tools while collaborating with engineering and research teams.

You will optimize distributed training, manage data versioning, and turn prototypes into repeatable pipelines that scale across GPUs. Strong Python/PyTorch skills are essential.

Qualifications

• 3+ years building ML systems or research infrastructure, including distributed training runs on 100+ GPUs.
• Expert-level Python and deep knowledge of PyTorch internals: DDP and FSDP, mixed precision, gradient accumulation, and profiling tools.
• Track record of building/maintaining research libraries others depend on (e.g., torch_geometric, torchaudio, torcheeg).
• Experience with data pipelines over large unstructured/multimodal datasets; familiarity with Parquet, Arrow, WebDataset, etc.
• Hands-on experiment tracking and dataset versioning tools (ClearML, W&B, MLflow).
• Ability to read papers, reimplement components, and assess loss curves.
• Debugging range from corrupted shards to silently wrong collate, etc.
• Care for well-being of people touched by this tech; thrive in high-bandwidth, collaborative env.

Responsibilities

• Orchestrate and optimize training: distributed configuration, mixed precision, checkpointing, and profiling.
• Build high-throughput data loading over TB-scale multimodal data: sharding, prefetching, caching.
• Wrangle complex data: align and synchronize multi-stream time series; dataset versioning and lineage.
• Build pipelines that derive rich features from raw recordings; versioning and reusable pipelines.
• Write well-documented libraries for datasets, models, transforms and metrics.
• Ensure runs are traceable by recording code version, config, dataset version, and environment.
• Create evaluation harnesses that run on new checkpoints and surface regressions.
• Turn research prototypes into repeatable pipelines without stifling flexibility.

Skills

Python

PyTorch internals

Distributed training

Data pipelines

Experiment tracking

Debugging

Collaboration

Research literacy

Tools

torch_geometric

torchaudio

torcheeg

torch_brain

neuralsets

Job description

Open role

Research Engineer

San Francisco (On-site), Full-time

About Us

Constellation is creating the AI-human translation layer that ensures humanity evolves alongside our technology. Our mission is to leverage AI towards addressing deep and meaningful problems at the core of the human experience: empowering people towards their goals, augmenting our cognition and emotional wellness, and understanding ourselves and each other. Our path forward is to move away from AI that captures human knowledge towards AI that truly understands what it is to be human. We are generating the richest multimodal dataset ever collected to build a new class of foundation models and we're seeking the team of researchers that will build them.

The Role

We're looking for a Research Engineer to sit between our data and our models and make the whole loop faster. You will orchestrate and optimize training runs on long-horizon multimodal sequences, build the pipelines that turn a messy, daily-growing corpus into something our models can learn from, and write the research code (libraries, dataloaders, evaluation harnesses) that lets the rest of the team try ideas quickly and trust what they see. You'll work closely with both our engineering and research teams.

Responsibilities

• Orchestrate and optimize training: distributed configuration (DDP/FSDP), mixed precision, checkpointing and recovery on multi-day runs, and profiling to find which of kernel, dataloader, communication or pipeline is actually the bottleneck before touching anything.

• Build high-throughput data loading over TB-scale multimodal data: sharding, prefetching, caching and format choices that keep GPUs saturated rather than waiting on I/O.

• Wrangle complex data: align and synchronize multi-stream time series, handle changing collection protocols, and own dataset versioning and lineage as the corpus grows.

• Build pipelines that derive rich features from raw recordings so every derived dataset is versioned, reproducible from its inputs, and cheap to recompute when upstream data changes.

• Write research code that makes research better: clean, well-tested, well-documented libraries for datasets, models, transforms and metrics that the team builds on rather than around.

• Make every run traceable: code version, config, dataset version and environment recoverable from any result.

• Build evaluation harnesses that run automatically on new checkpoints and surface regressions before anyone goes looking.

• Turn research prototypes into repeatable pipelines without flattening the flexibility researchers need to keep exploring.

Qualifications

• 3+ years building ML systems or research infrastructure, including distributed training runs on 100+ GPUs.

• Expert-level Python and deep knowledge of PyTorch internals: DDP and FSDP, mixed precision, gradient accumulation, and the profiling tools to tell a slow model from a starved one.

• A track record of building or maintaining research libraries others depend on. Contributions to packages like torch_geometric, torchaudio, torcheeg, torch_brain, neuralsets, or comparable internal tooling are exactly what we're looking for.

• Experience with data pipelines over large unstructured and multimodal datasets, and familiarity with columnar and streaming formats (Zarr, Parquet, Arrow, Lance, WebDataset, Vortex) and the tradeoffs between random access and sequential throughput.

• Hands-on experience with experiment tracking and dataset versioning tooling (ClearML, Weights & Biases, MLflow, or similar).

• Enough research fluency to read a paper, reimplement a component, and tell whether a loss curve is broken.

• Debugging range from a corrupted shard to a silently wrong collate function to a training run that quietly diverged on day two.

• You care about the well-being of the people this technology touches, and you thrive in a high-bandwidth, collaborative environment.

Nice to Have

• Custom kernel work in CUDA or Triton, or compiler-level optimization (torch.compile, TensorRT, ONNX).

• Experience with time-series or multimodal data where synchronization across streams is a prerequisite for analysis and learning.

• Low-latency or edge inference, quantization, or distillation.

• Comfort dropping into Rust or C++ when Python is the wrong tool.

• Interest in neurotech, mental health, or human-AI interaction.

Benefits

• Comprehensive, high-quality health, dental, and vision insurance with premiums fully covered.

• A renovated, light-filled office with a quirky layout and full kitchen in the heart of San Francisco's Mission District, surrounded by world-class food and coffee.

• Relocation assistance for those joining us in SF and workspace setup stipend.

• Bi-monthly team dinners and outings, plus twice-yearly off-sites and retreats.

• Fully stocked kitchen (snacks!) and a dedicated meal budget.

• Sponsorship for travel to relevant conferences and regular professional development activities to help you level up.

• A minimum of 12 weeks of fully paid parental leave with a "soft-landing" transition back.

• Paid time off (PTO)

Compensation

$180K - $250K Offers Equity