SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Twelve Labs is building the intelligence layer to understand video at scale. Video represents 90% of the world's data, yet most remains invisible to machines. The company's multimodal AI models understand video across sight, sound, and motion, powering production-scale AI workloads in media, entertainment, sports, security, and government. Backed by $210M+ from NEA, Radical Ventures, Amazon, NVIDIA, Snowflake, and others, Twelve Labs operates globally from San Francisco with offices in Seoul, New York, and London.
Jockey is Twelve Labs' unified agentic system that reasons across videos and images, combining a reasoning model with a memory layer to build knowledge stores from large video corpora. Unlike traditional context-window approaches, Jockey decomposes queries, retrieves, segments, and reasons across thousands of videos and images—handling archives of millions of hours. Jockey Core is the reasoning LLM at the center of this system, responsible for decomposing queries, deciding what to retrieve and segment, and reasoning over results into actionable answers.
The Cognition Models team owns the models that turn video into structured understanding and reasoning, including Pegasus (video-language model) and Jockey Core. The team focuses on multimodal systems with high instruction-following capability and complex hierarchical outputs, spanning training infrastructure, temporal segmentation, structured metadata extraction, large-scale inference and serving systems, data curation, and evaluation pipelines.
In this Tech Lead Manager role, you will build and lead a newly founded team developing Jockey Core. You'll own the end-to-end roadmap from model and engine selection through model-efficiency work (pruning, quantization, distillation) to production serving and scale-out. You'll stay deeply hands-on while standing up and growing the team, leading critical system and serving/inference architecture decisions, setting technical bar through design review, and judging latency/throughput/cost tradeoffs with measured data. You'll partner with Pegasus, agent, and infrastructure teams on capacity and SLOs, and explore AI-assisted development tools to raise team productivity.
Ideal candidates have a track record leading ML/infrastructure teams as a hands-on tech lead or manager, ideally founding or scaling teams from small. Deep experience serving and optimizing large-scale LLM inference in production (vLLM, TensorRT-LLM, SGLang) across batching/scheduling, quantization, disaggregated prefill/decode, and speculative decoding is essential. You should drive ambiguous technical decisions with measured data and demonstrate excellent communication and people leadership.
