Python Tech Lead / Integration Architect (Synthetic Data Generation)

Description

Description

As a Tech Lead / Integration Architect for the Nebula project, you will lead the technical design and implementation of a synthetic railway data visualization platform for the Customer. This Python-heavy architecture role focuses on building a video processing pipeline that converts NVIDIA Omniverse simulation outputs (Camera + LiDAR sensors) into synchronized HLS video streams delivered via web player. The critical technical challenge is maintaining ≤100ms synchronization drift between dual video streams throughout the entire timeline. You will architect the system from scratch, implement core Python components, and guide a 3-person engineering team to deliver a production-ready solution by March 31, 2026. This is a greenfield project with high technical ownership, involving FFmpeg automation, multi-sensor synchronization, file-based system integration with Thanos simulation environment, and REST API design for Angular frontend consumption.

Requirements

Qualifications

• 7+ years of professional Python development experience
• Strong understanding of system architecture and pipeline design
• Strong programming skills in Python 3.11+, asyncio, multiprocessing, advanced patterns
• Experience with FFmpeg integration and video processing via Python
• Experience with HLS/DASH streaming protocols (manifest generation, segmentation, packaging)
• Experience with video codecs (H.264/H.265) and multi-stream synchronization
• Experience with REST API development using FastAPI or Flask
• Strong file parsing expertise (CSV/JSON/binary formats, large file handling)
• Experience with Docker containerization and Linux/Unix systems
• Proven technical leadership experience (2-3 engineers minimum)
• Excellent communication and presentation skills

Nice to have

• USD format knowledge (Universal Scene Description) and Python USD libraries (pxr or omni.usd)
• Experience with NVIDIA Omniverse, Houdini, Unreal, Unity, or similar 3D platforms
• LiDAR point cloud data formats and sensor fusion concepts
• Experience with CARLA, AirSim, or similar simulation platforms
• Automotive, robotics, or railway industry background

• NumPy/Pandas for data processing, OpenCV for video manipulation

   

Job responsibilities

Description

As a Tech Lead for the SimTrack project, you will architect and lead the development of a Python-based video processing pipeline that synchronizes dual-sensor (Camera + LiDAR) railway simulation outputs into HLS streaming format. You will design the end-to-end system from file parsing to web delivery, ensuring ≤100ms synchronization between video streams, while guiding a small engineering team to deliver a production-ready solution for Cuastomer’s AI training datasets.

Key Responsibilities

• Design system architecture: Input files → Python processing pipeline → HLS streams → Web player integration
• Build video encoding pipeline using FFmpeg and Python automation for Camera + LiDAR streams
• Develop file parser for odometry log processing (speed, position, orientation, timestamp data)
• Implement scene file processing to extract sensor data and metadata from simulation outputs
• Create HLS packaging automation ensuring ≤100ms drift between Camera and LiDAR timelines
• Architect and implement metadata synchronization engine (odometry data → video timeline mapping)
• Define and implement REST API contracts for frontend integration using FastAPI
• Guide and code review Backend Engineer (Python) and collaborate with Frontend Engineer (Angular)
• Ensure file-based integration (S2L format) with Thanos simulation environment works correctly
• Collaborate with 3D Artists on file format requirements and with R&D consultant on sensor validation
• Perform performance optimization, bottleneck identification, and system troubleshooting
• Document architecture, technical decisions, API specifications, and integration guides
• Lead technical problem-solving and make critical architectural decisions under tight timeline