Voyager Wingman

I have filed a few issues on Github. The first one is a week old and hasn’t been looked at. I’ve uncovered a few plumbing issues today as well. Most of what Wingman ships is already readable: the VS Code base and the bundled Codex engine are open source, and the wingman-dashboard extension ships as unminified JavaScript. Publishing the source here would cost little and would help in a few concrete ways:Contributions instead of reports. Several of the open issues are one-line fixes. With the source here, they could arrive as pull requests. Packaging for more distributions. Packagers for Arch, Fedora and others could build Wingman properly instead of repackaging the .deb. Trust. For a tool that runs an agent with shell access on developers' machines, being able to review the code is a strong selling point, especially for a European company competing with much larger vendors.If publishing the whole codebase isn't an option, opening just the Linux launcher (/usr/bin/wingman) and the packag

Hello Axelera community,I wanted to share a user story of how Wingman allowed me to increased the performance of my benchmarking harness from 695 FPS to 1601 FPS. Same AI module, same model. One missing property.Edge AI Coding — Part 3: Axelera Wingman | Mario BergeronIf you are wondering what Axelera’s Wingman is bringing to the table, compared to general-purpose coding assistants, these are my 2 cents.I love where Axelera is going :)Cheers !

Sharing a measurement in case it saves someone else the confusion, and because it changes how you would size a deployment.Setup: Metis M.2 (metis-0:3:0, firmware 1.7.0, bc 7.4, 800 MHz), Voyager SDK 1.7.0, driver metis-dkms 1.5.5, yolov8n-coco-onnx, --aipu-cores 4 --pipe gst, Ubuntu 24.04.┌─────────────────────────────────────────────┬───────────┬───────────┐│ │ 1080p │ 4K │├─────────────────────────────────────────────┼───────────┼───────────┤│ inference.py … --output out.mp4 (16 clips) │ 34–61 FPS │ 13–16 FPS │├─────────────────────────────────────────────┼───────────┼───────────┤│ Headless app on create_inference_stream() │ 740 FPS │ 238 FPS │├─────────────────────────────────────────────┼───────────┼───────────┤│ SDK stage table, Metis element (both cases) │ ~840 FPS │ ~840 FPS │└─────────────────────────────────────────────┴───────────┴───────────┘The headless 1080p figure (740) sits just under the AIPU stage rate (~8

I was watching the new Wingman walkthrough from our agentic coding whisperer ​@Abhilash and I was expecting a "build from a prompt" demo, and while there is plenty of that in there, but the bit that realy leapt out at me is about porting.If you've already got a computer vision project running on NVIDIA or Hailo or similar, you don't actually have to rebuild it from scratch to move it over. You open the Wingman app, hit "Port to Axelera", and it takes over from there. Wingman creates the Voyager pipeline, adapts your existing application logic, compiles it, runs it, validates it on the hardware, and bakes a celebration cake for you afterwards.What started life as a project on another platform is suddenly running on Axelera, with the result sitting on the dashboard ready to inspect and refine. One click.For anyone who's ever done a hardware migration by hand, you know why that's a big deal. Usually that's the part that eats weeks. Rewriting the pipeline, chasing model compatibility, re-t

Hi Axelera Community After some prompts in the Axelera Wingman browser version, I wanted to test on the Metis M.2 hardware installed in an IPC on my desk.After downloading and installing the Linux amd64 client (version 1.4.0), logging in is required. Where the browser version supports logging in with an Axelera Community account or with Axelera SSO, the desktop client only seems to show the Axelera SSO option, which is not possible for me of course. Am I missing anything here? How should a non Axelera employee login to the desktop app? Kind regardsRobbe

Did you know that wingman learns from your preferences and projects?   What’s your favourite feature?