Weekly Round Up | OpenAI DevDay, ElevenLabs Workflows & Gemini’s Nano Banana

OpenAI’s DevDay turns ChatGPT into an app platform, ElevenLabs launches Agent Workflows for smarter conversations, and Google’s Nano Banana shows 4 ways AI is reshaping photo editing.

Author

Adam Stewart
October 09, 2025

Adam here! This week, AI is evolving from smart assistants into full ecosystems built for creators, developers, and teams.

OpenAI’s DevDay 2025 turned ChatGPT into a full app platform, complete with its own store, tools, and a glimpse at AI-first hardware.

ElevenLabs unveiled Agent Workflows, a visual way to build structured multi-agent conversations that blend automation with human logic.

And Google’s Nano Banana in the Gemini app shows how photo editing is becoming conversational, no pro tools, just prompts.

Here’s what’s shaping how we’ll build, connect, and create with AI next.

OPEN AI

DevDay 2025: ChatGPT Becomes an App Platform

Open AI: Sam Altman Opening Keynote at DevDay

OpenAI’s DevDay 2025 introduced the next evolution of ChatGPT. From a chat assistant to a full AI app platform where developers can build and sell apps directly inside the interface.

  • ChatGPT Apps: Launch apps like Spotify, Canva, or Figma straight inside ChatGPT.

  • AgentKit: New developer tools for building autonomous AI agents.

  • App Store + Monetisation: OpenAI is rolling out a directory and revenue model for creators.

  • Hardware Vision: Teased collaboration with Jony Ive for future AI-first devices.

ChatGPT is becoming more like an operating system for AI where browsing, building, and business all happen in one smart environment.

AI AGENTS

ElevenLabs Launches Agent Workflows: Build Smarter Conversation Flows

ElevenLabs just introduced Agent Workflows, a visual editor inside its Agents Platform that lets you design structured conversation flows — no complex coding required.

  • Visual logic builder: Create and connect conversation paths in a drag-and-drop interface.

  • Subagents: Route complex requests to specialised subagents, each with its own system prompt, tools, and knowledge base.

  • Human handoff: Define when to escalate conversations to a live operator for a seamless experience.

  • Efficiency boost: Optimise cost and speed by assigning the right LLM for each step in a workflow.

Agent Workflows bring enterprise-grade structure to conversational AI — giving teams fine control over logic, security, and scalability without writing massive prompt trees.

AI IMAGING

4 Tips for Using Nano Banana to Create Better Images in Gemini

Google DeepMind’s Nano Banana model in the Gemini app keeps evolving, and creators are still discovering clever ways to use it. A new post from Google highlights four simple but powerful tips:

  • Stay consistent: Keep the same character or scene across edits. Change outfits, angles, or lighting while keeping the likeness intact.

  • Edit with precision: Use natural language for “pixel-perfect” tweaks, like changing a sign’s text or recolouring a sofa without touching the rest of the image.

  • Get creative: Combine up to three images to remix scenes or restore old photos with colour and context.

  • Build fast: Try PictureMe in Gemini Canvas — an app built on Nano Banana that transforms one photo into portraits across decades or styles.

Nano Banana shows how multimodal AI is making advanced photo editing conversational + no pro tools, just prompts.

TWEETS OF THE WEEK
MORE AI IN ACTION
DISCOVER THESE AI TOOLS AND APPS

  1. Klap AI — Create TikToks, Reels, Shorts from your long videos in just one click.

  2. Gamma + Zapier Automation — Turn Every Zoom Call Into a Professional Presentation

  3. AgentKit by OpenAI — a complete set of tools for developers and enterprises to build, deploy, and optimize agents.

  4. Apps in ChatGPT — A new generation of apps you can chat with and the tools for developers to build them.

  5. GPT5 Codex — OpenAI's series of AI coding tools that help developers move faster by delegating tasks to powerful cloud and local coding agents.

  6. Sora 2 — OpenAI's advanced AI video generation model that creates high-quality, realistic videos from text descriptions.

  7. Gemini 2.5 Computer Use model — Available in preview via the API, our Computer Use model is a specialized model built on Gemini 2.5 Pro’s capabilities to power agents that can interact with user interfaces.

  8. Grok Imagine — Grok Introduces their new video generation model with massive upgrades from v0.1 in visual quality, motion, audio generation, and more.

  9. EbSynth 2 — is a VFX software for creative video transformations, retouching, and rotoscopy.

  10. Spotify Personalisation using ChatGPT — Bring Spotify into their ChatGPT conversations to receive personalized music and podcast recommendations.

Curious about AI? So are we. Join the AI Systems Lab community and explore what’s next, one experiment at a time → aisystemslab.ai