Responsibilities
- Build agentic systems that can reason, plan, and execute multi-step tasks reliably
- Design and deploy LLM-based services (tool use, structured outputs, evals, monitoring)
- Develop video/image intelligence pipelines (retrieval, matching, classification, metadata extraction, generative augmentation)
- Integrate existing APIs and tools when appropriate (strong build-vs-buy judgment)
- Ship scalable inference pipelines on cloud infrastructure
- Apply MLOps / LLMOps best practices: observability, testing, evals, versioning, cost control
- Work closely with founders and product to iterate quickly based on user feedback
Requirements
- Strong Python engineering skills and proven ability to ship production systems
- Experience building with LLMs (OpenAI-style APIs, tool calling, RAG, structured generation)
- Experience deploying ML/LLM services in production (cloud, containers, serverless, async pipelines)
- Practical understanding of video/image processing (CV or media pipelines)
- High autonomy: you can take vague problems and turn them into shipped solutions
- Strong engineering fundamentals: clean code, reliability, performance, and cost awareness
Nice to Have
- Agent orchestration frameworks (LangGraph, LangChain, AutoGen, CrewAI, etc.)
- Experience with vector search / embeddings / retrieval systems
- Experience with workflow orchestration (queues, batch jobs, event-driven systems)
- Experience with PyTorch and GPU inference optimization
- Experience working on media-heavy products (video editing, video generation, large file pipelines)
