AWS Reveals Details About Amazon Q at NY Summit

AWS Summit NY took place July 10, 2024 at the Javits Center in NYC. I’ve been to this event before and it’s always fun. There were a few protesters out front like last time. But the vibes were good as a few thousand people streamed in, walked the show floor, attended “Silent Sessions” (four adjacent stages with everyone wearing headphones) or smaller sessions in side rooms. The big difference with 2023 was the laser focus AWS placed on AI and especially the rollout of their “Q” tools. Here’s a recap of what I learned about Q.

“Customers absolutes love Sagemaker” … talking about Sagemaker Studio … Q now being added to Sagemaker Studio

[PICTURE 1]

Q will find errors in Sagemaker code …

[PICTURE 2]

Widest selection of models … need to match model to mission …

[PICTURE 3]

Model evaluation (find the right model)

Fine tuning …

[PICTURE 4]

Training models on private corp data …. (Retrieval Augmentation)

[PICTURE 5]

Knowledge Bases (new data connectors) …

[PICTURE 6]

Models + RAG” … Guardrails …

[PICTURE 7]

New ways to avoid hallucinations …

[PICTURE 8]

Guardrails API …

Put AI into action with “Agents” …

Multi step tasks

LLM for analytics … use to generate “code” to run analytics …

All this was leading up to the big reveal around Amazon Q.

Q for clean code dev …

Use for code transformation (ie translate diff versions of Java)

Q Developer Agent

Example: Store and retrieve recipes from DynamoDB

Q learns your own code, APIs etc. so it can give your code recommendations based on your unique code base

Q apps creator …

As an example they show Bayer’s food science division. They need to analyze a huge amound of crop drone video and build models around that.

New. AWS App Studio …

Related Posts