Act as the Team Lead for the Evidence Domain, overseeing sprint planning, grooming, and driving the team toward critical delivery milestones.
Propose and defend robust architectures for high-stakes problems, leveraging GKE, Pub/Sub, Vertex AI, and Cloud Storage to build fault-tolerant systems.
Architect scalable media processing pipelines, including distributed transcoding, storage optimization for high-volume video, and digital integrity verification (SHA-256) to ensure court-ready evidence.
Architect a vendor-agnostic "Adapter Pattern" for evidence transfers, ensuring the platform can securely handshake with any third-party repository via OAuth 2.0 and mTLS.
Partner with Product Management to translate broad visions into actionable technical stories and architectural blueprints.
Drive interactions across infrastructure and security teams to resolve dependencies and ensure the platform remains cohesive and performant.
Lead the implementation of LLM-powered features (e.g., automated evidence summarization via Vertex AI), ensuring rigorous monitoring of live metrics and feedback-loop integration.
Serve as a subject matter expert (SME) across the organization, coaching engineers and setting the standard for technical excellence.
Requirements
Bachelor’s degree with 5+ years of full stack engineering and Python experience AND 3+ years of Java experience
Legal authorization to work in the U.S. indefinitely is required.
Rust and C++ experience is preferred.
Recognized expert in the Google Cloud ecosystem, specifically GKE, VPC Service Controls, and IAM Workload Identity.
Proven experience architecting large-scale video processing workflows, including worker-pool strategies for transcoding, bitstream manipulation, and metadata preservation.
Deep experience with event-driven architectures, state machine design, and managing complex service-to-service contracts.
Proven ability to lead through influence, facilitate technical consensus among senior peers, and drive team velocity.
Demonstrable experience using modern AI tools (e.g., Claude Code, Cursor) to accelerate the SDLC and automate complex task breakdowns.