SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Eventual is building infrastructure for Physical AI and robotics teams to find, curate, and stream multimodal data (video, LiDAR, radar, sensor telemetry) at petabyte scale. Founded in 2022 and backed by $30M from Felicis, CRV, Y Combinator, and Databricks/Perplexity co-founders, Eventual's open-source engine Daft powers data pipelines at Amazon, FAANG companies, Mobileye, and TogetherAI.
As a Software Engineer on Multimodal Backend Systems, you will own core product architecture and ship features used immediately by customers. The role spans three primary domains:
**Real-Time Video Infrastructure**: Design and optimize WebRTC media pipelines and custom signaling for multi-camera robot video feeds. Integrate hardware-accelerated video codecs (H.264, HEVC, AV1) using GPU acceleration (NVENC/NVDEC) with dynamic bitrate adaptation. Engineer high-throughput video ingestion and distributed transcoding services that compress, index, and write petabyte-scale media to AWS S3/GCS.
**Robotics & Physical AI Telemetry**: Build specialized storage formats and spatial indexes for dense 3D LiDAR point clouds, depth maps, and multi-camera datasets. Develop high-throughput ingestion pipelines for real-time sensor streams and robot state telemetry across customer fleets. Implement time-sync protocols (PTP/NTP) to align LiDAR, IMU, and video data into unified structures.
**Product & Multimedia Frontend**: Architect high-performance web applications using React and TypeScript for continuous data streams. Build custom frontends to render 3D LiDAR point clouds, spatial bounding boxes, and multi-camera video. Create agentic data-exploration and search tools for robotics operators and AI researchers.
You'll work with an experienced team from AWS, Lyft, and Tesla in a tight-knit environment. The role demands strong fundamentals in Rust, C++, Python, or Go, with willingness to reach for lower-level primitives when performance demands it. You'll have high ownership across compute infrastructure, data storage/querying, and model training/deployment. The team operates 4 days/week in-office at the SF Mission District office.
