On Wednesday, we'll cover the infra stack that we recommend for RAG in 2025. Then, we'll build, ship, and share a best-practice RAG app.
β
We'll also discuss important production tradeoffs and implications that you should consider before and after deployment when going from zero to production RAG!
β
Last week, we discussed the latest open-source repo drops from DeepSeek Week, and we covered how they're being used as a new best-practice way to do inference on MoE models via Hopper (e.g., H100, H200, etc.) GPUs! vLLM has already implemented much of the inference system, as we saw during the event! Learn how cross-node Expert Parallelism drove the solution to the problem the DeepSeek team sought to solve of higher throughput and lower latency, resulting in decreased cost and high theoretical income.
This is why DeepSeekβs inference system has been called βThe North Star for LLM Inferenceβ by vLLM
π Coming Up!
β
Enterprise Agents with OpenAI
What does the agents SDK look like from OpenAI? How does it build on previous work they've done? Are they officially in the end-to-end platform game competing with orchestration frameworks like LangChain, LlamaIndex, CrewAI, and others? Join us live to find out!
What is MCP, exactly? What can we learn about looking at the protocol as defined? What can we learn from tracing its impact since rits elease over the past 5 months? Released by Anthropic in November 2024, MCP has caught on as a new gold standard for the way that context is shared between models. Let's learn it, then build, ship, and share with it!
π‘ Transformation Spotlight: Tyler Laughlin, an Information Systems Engineer at Adobe. He has some great advice for enterprise-level engineers on how they should be using Gen AI. Read more about his story!β