Kotlin Weekly #513 has just been published! - mailchi.mp/kotlinweekly...

Adding embeddings/RAG support to the Koog-based AI agent in Confetti johnoreilly.dev/posts/confet... This is using Gemini for the LLM and the embedding model so might as well use it to generate image for the article 😀

Agenda for @droidcon.bsky.social USA 2026 now included in Confetti (github.com/joreilly/Con...)

Is there a word for that feeling when you're out and about and you forgot to enable remote control on Claude back at your computer? 😄

🧩 How do you unify Android and iOS code without rewriting everything? In this case study, Workday shares how they adopted Kotlin Multiplatform step by step to reduce duplication, share business logic across platforms, and fit KMP into native apps. Read the case study 👉 https://kotl.in/c58mnc

Starting to add a Koog AI agent to the Confetti KMP sample (github.com/joreilly/Con...). The project had been using Gemini directly before for providing recommendations but having an agent like this (along with associated tools) is a lot more flexible (still using Gemini as LLM)!

I've been so focussed on @kotlinconf.com talk for last few months I've been somewhat neglecting my KMP/CMP samples (github.com/joreilly). Waiting at an airport is perfect time to start updating them again!

I took very few pictures at @kotlinconf.com, but this one is a goodie. With @johnoreilly.dev and @russhwolf.com at the day 1 keynote. Two legends of Kotlin OSS. Thanks for the chat!

This time 40 years ago I was getting ready to do my Leaving Cert (final school exams in Ireland). I was so focused on doing Electronic Engineering in NIHE (now UL) that it was only college option I'd selected! A few months later I'd be learning C and knew programming was my future!

Live coding talk at @kotlinconf.com about using Koog in a Compose Multiplatform application completed ✅. Very happy also that I was able to demo new capability to use local models on Android that was released yesterday! (thanks @benoitdion.bsky.social for photo)

Very cool to see LiteRT-LM integration working now with Koog 1.0. That version also brings OpenTelemetry to other KMP targets. Following is example of telemetry in langfuse for a Koog agent using Gemma 4 E2B model running on an Android device (including tool usage)

KotlinConf'26 started today in Munich, and there’s a lot to announce. Kotlin is easier than ever to get started with, and it now powers some of the most-used apps in the world – both on the backend and mobile. 🧵 Here are the highlights from the opening keynote.

All set for @kotlinconf.com tomorrow!

An Irish person in the English Gardens in Munich practicing their @kotlinconf.com talk 😃. Getting grounded by nature in the same way I'll be showing how AI agents are grounded by tools! Sorry, couldn't resist 😃

Android is now Compose-first: developer.android.com/develop/ui/c... It was years in the making, and I'm happy to see it's finally here!

A New Default Project Structure for Kotlin Multiplatform blog.jetbrains.com/kotlin/2026/...

I'm having issues with my knee at the moment and got MRI scan today. Need to wait a few days for formal analysis but in meantime they gave me access to images (all 154 of them!).....which of course I uploaded to Claude to analyse! This is small subset of its findings!

Confirming I can if needed work offline for my live coding talk at @kotlinconf.com. Have local Ktor server emulating endpoint and also local ollama model.....and appropriately enough thought I'd try following to make some updates! ollama launch claude --model qwen3.5:35b-a3b-coding-nvfp4

Following is model that seems to perform best on MacBook here atm....running as part of a Koog agent (showing telemetry using @langfuse.bsky.social). Full agent run takes about 2s with LLM calls taking (in this case) 1.52s and 0.31s. Memory footprint ~20GB. qwen3.5:35b-a3b-coding-nvfp4

💡 Getting started with Kotlin Multiplatform? Explore official learning resources with tutorials, video courses, articles, and documentation – all in one place. From the basics to advanced topics, learn KMP at your own pace. Start here 👉 https://kotl.in/c0zzvf

compose-ai-tools - Render @Preview composables to PNG outside Android Studio, so AI coding agents can see what they're changing. github.com/yschimke/com...

Say hello to Compose Hot Reload 1.1 🔥 Please give it a try and provide feedback: github.com/JetBrains/co... ☕️ JBR 25 support ⚡️ Recompilation Speed optimizations 😎 New visuals, New DevTools UX 📦 Faster, more lightweight protocol under the hood

Claude is so useful for stuff like this!

Been a pretty good week in the West of Ireland! ☀️

New version of llmfit (github.com/AlexsJones/l...) today that supports Qwen3.6 models

Bring pets to work day today 😃

Still find myself fixing typos and adding question marks (when clearly a question) in Claude etc prompts!

Managing a series of cascading PRs with Claude is so useful....including monitoring CI, fixing issues found, rebasing the various branches and pushing/retrying again. And all from my phone thanks to /remote-control.

Introducing Claude Opus 4.7 www.anthropic.com/news/claude-...

AI-focused sessions are now live for KotlinConf’26! Join experts exploring scalable agents high-quality AI development, and Kotlin-powered machine learning across platforms. Browse the full schedule and stay ahead of the curve! ➡️ https://kotl.in/conf26-schedu

Trying out Claude Code + local Ollama model for something a bit more involved ("support viewing information from api.climatetrace.org/v7/docs/inde...) ....did a pretty good job! Started with empty Compose Multiplatform project. ollama launch claude --model qwen3.5:35b-a3b-coding-nvfp4

Still trying local LLMs on M5 Max device here. Was comparing following and was interesting to see explanation from Gemini below about why Qwen model is faster (and certainly seeing that's the case here in testing). - gemma4:26b-a4b-it-q4_K_M - qwen3.5:35b-a3b-coding-nvfp4

llmfit (github.com/AlexsJones/l...) is nice way to see what models run best on your hardware

Confetti now has the AndroidMakers data thanks to @zsmb.co 🙏 confetti-app.dev

M5 Max laptop arrived and starting to try out various local models (as part of Koog AI agent using Ollama integration). Latency numbers looking pretty good so far. This one is a 4B parameter model but working really well for example I'm using (e.g. very reliable tool calling)

The world of #AI is expansive and ever changing. Most of it filters down for all practical purposes to a couple of key concepts and tools. It's more than sufficient to know those and expand your knowledge if/when needed.

Interesting Fragmented Podcast episode on Background AI Agents fragmentedpodcast.com/episodes/309/

Another useful Claude prompt! "add comments to all the Kotlin files....write in way to use as "speaker notes" during live coding session"

Very cool to see! Showing this data in Claude Desktop using MCP Server from the ClimateTraceKMP sample (github.com/joreilly/Cli...). That makes use of the Kotlin MCP SDK (github.com/modelcontext...) and Ktor to connect to the @climatetrace.org endpoint.

Ollama 0.20.0-rc0 available and with it the new Google DeepMind Gemma 4 local model. Trying it here as part of a Koog AI agent (screenshot thanks to Koog's OpenTelemetry langfuse integration)

And then there were 14! (from github.com/joreilly/Peo... #KMP/#CMP sample)

This is either brilliant or scary: Anthropic accidentally leaked the TS source code of Claude Code (which is closed source). Repos sharing the source are taken down with DMCA. BUT this repo rewrote the code using Python, and so it violates no copyright & cannot be taken down!

Tried new qwen3.5:4b-nvfp4 Ollama model on M1 Max here (in project where it's used with Koog AI agent)...38% faster than qwen3.5:4b (averaged over 5 runs of the agent). I asked Gemini and it initially claimed that I shouldn't see increase in performance like this...asked again and got following 😀

The latest 2.4.0 dev build of Swift Export now supports exposing StateFlow directly to Swift code. Chip-8 sample updated to use it. #KMP github.com/joreilly/chi...

KotlinConf 2026 is packed with inspiring talks in multiple tracks! Not sure how to plan your agenda? Our new blog post highlights a few sessions you might want to check out. 👀 https://blog.jetbrains.com/kotlin/2026/03/kotlinconf-2026-talks-schedule/

Compose Hot..Swan is finally here! 🦢 Compose HotSwan enables Compose Hot Reload in real Android devices, instant UI updates on your running Android app, with state preserved. No rebuilds, no relaunching, no navigating back to the same screen again. hotswan.dev

Loving the Claude/Slack integration.....so nice to be able to use for querying for example history of discussion on particular topic or conclusions reached etc

Good news from kotlinx-rpc! They now have a dev preview available where you can use gRPC with Kotlin Multiplatform! Check out the release notes and resources to try it:

The KotlinConf’26 schedule is live. ✨ Discover upcoming talks on Kotlin’s evolution, AI-powered development, and real-world data workflows. Join us to learn, connect, and grow! ➡️ https://kotl.in/conf26-schedu