AI deployment and orchestration research

You will be researching deployment strategies for LLM powered agents. The research will be published in video format on YouTube. For reference see this research, https://careersatdoordash.com/blog/how-we-learned-to-trust-our-ai-code-reviewer-at-doordash/ 1. Framework & Harness Evaluation: Install, configure, and systematically test various open-source and commercial LLM deployment frameworks, agent harnesses (e.g., LangGraph, AutoGen, CrewAI), and post-training/fine-tuning methodologies. 2. Cost & Performance Benchmarking: Design and execute benchmark tests across a mix of small, medium, and large language models to measure inference latency, token throughput, computational efficiency, accuracy, and total operational cost per use case. 3. Enterprise Use Case Validation: Map benchmarking insights to practical, cost-effective enterprise and small-business workflows, verifying how different framework-model combinations handle real-world tasks. 4. Technical Documentation & Research Writing: Maintain meticulous documentation of test environments, prompts, configurations, and raw data, translating these findings into clear, publication-ready research articles. 5. Video Content Production: Produce, record (screen capture/walkthroughs), and assist in editing high-quality, long-form YouTube videos that clearly communicate the research methodologies, code implementations, and benchmarking results to the global developer community. 6. Code Repository Management: Organize, clean, and maintain public or internal GitHub repositories containing the benchmark code, configuration files, and reproduction steps for the community to verify the published research. We aim to create 50 such videos. For each video that is approved and published, we will pay $40.