The Cloud Pod | Weekly AI & Cloud News on AWS, Azure & GCP · Justin Brodley, Jonathan Baker, Ryan Lucas and Matt Kohn | Cloud Computing & AI News

Azure releases a new SQL AI Assistant… Jimmy Droptables

·1 hr 13 min·4 clips

Gemini 3 Flash can now zoom, crop, rotate, and annotate images through Python execution.

The panel opens with the usual CloudPod rhythm: somebody is missing, somebody gets blamed, and the notes are treated like a loose suggestion. Matt is out at a dinner for his day job, and the hosts waste no time enjoying the fact that he is not there to defend the Azure stories. Justin says he is doing better now that half the topics have been cut. Ryan and Jonathan immediately lean into that joke, which is basically the show’s native language. The headline Azure item is a new SQL AI assistant called Jimmy Droptables. The hosts do not spend long on it. That is part of the joke too. They care more about whether any cloud news is actually worth the air it takes to explain. A brief detour into SAS gets flattened into something less dramatic than the name suggests. That fits the tone of the room: technical, but unwilling to pretend every acronym deserves a standing ovation. After Azure, the conversation shifts to Google. Google has partnered with deeplearning.ai on a free, comprehensive course for Gemini CLI. The tool is framed as an open-source command-line agent that folds AI into daily workflow without much ceremony. The hosts like that the course is short enough to finish in under two hours and organized into nine lessons. It moves from basic setup into more specific workflows. The point is not just that the course exists. It gives people a practical way into command-line agents without turning the whole thing into a weekend project. The panel spends time on context management through Gemini MD files. They also note the MCP protocol and how external tools can be wired in through MCP servers. Memory across sessions matters here, because the whole pitch depends on the tool keeping enough context to stay useful. Custom extensions fit into that same picture. The course is not only for developers. The hosts point out that it also covers data visualization from local CSVs and Google Sheets. It includes automated blog and social media content generation too. Another module helps users build study plans. That wider scope is what makes the course interesting. Google is trying to make the command line feel less closed off, and the show treats that as practical rather than flashy. There is some overlap with the hosts’ own interest in tools that cut down on grunt work. Ryan mentions reading about people using cloud code for spreadsheets and forms, which lands in the same neighborhood even if the thread runs out. The episode keeps its usual pace: quick banter, a compressed news read, then a practical step back to ask what the tooling means in real work. It ends with travel plans, a Super Bowl timing check, and the normal sign-off rhythm.

As heard by us

Cloud banter, a quick Azure swipe, and a practical Gemini CLI overview.

The episode starts in a loose cloud-news mode, with the hosts cracking jokes about vendor chatter, including a quick jab at Azure and an easy willingness to move past anything that does not hold their interest.

Read the full review in PlayNext →

Why you'd press play

Want the quick Azure joke and the useful Gemini CLI rundown?

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from The Cloud Pod | Weekly AI & Cloud News on AWS, Azure & GCP